Ethical implications related to processing of personal data and artificial intelligence in humanitarian crises: A scoping review

Ethical implications related to processing of personal data and artificial intelligence in humanitarian crises: A scoping review | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Ethical implications related to processing of personal data and artificial intelligence in humanitarian crises: A scoping review Tino Kreutzer, James Orbinski, Lora Appel, Aijun An, Patrick Vinck This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4224535/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 15 Apr, 2025 Read the published version in BMC Medical Ethics → Version 1 posted 10 You are reading this latest preprint version Abstract Background Humanitarian organizations are rapidly expanding their use of data in the pursuit of operational gains in effectiveness and efficiency. Ethical risks, particularly from artificial intelligence (AI) data processing, are increasingly recognized yet inadequately addressed by current humanitarian data protection guidelines. This study reports on a scoping review that maps the range of ethical issues that have been raised in the academic literature regarding data processing of people affected by humanitarian crises. Methods We systematically searched databases to identify peer-reviewed studies published since 2010. Data and findings were standardized, grouping ethical issues into the value categories of autonomy, beneficence, non-maleficence, and justice. The study protocol followed Arksey and O’Malley’s approach and PRISMA reporting guidelines. Results We identified 8,387 unique records and retained 98 relevant studies. One in four (n = 25) discussed technologies related to artificial intelligence. Seven studies included an author from a lower-middle income country while none included an author from a low-income country. We identified 22 ethical issues which were then grouped along the four ethical value categories of autonomy, beneficence, non-maleficence, and justice. Slightly over half of included studies (n = 52) identified ethical issues based on real-world examples. The most-cited ethical issue (n = 74) was a concern for privacy in cases where personal or sensitive data might be inadvertently shared with third parties. The technologies most frequently discussed in these studies included social media, crowdsourcing, and mapping tools. Conclusions Studies highlight significant concerns that data processing in humanitarian contexts can cause additional harm, may not provide direct benefits, may limit affected populations’ autonomy, and can lead to the unfair distribution of scarce resources. The anticipated increase in AI tool deployment for humanitarian assistance amplifies these concerns. Urgent development of specific, comprehensive guidelines, training, and auditing methods are required to address these ethical challenges. Moreover, empirical research from low and middle-income countries, disproportionally affected by humanitarian crises, is vital to ensure inclusive and diverse perspectives. This research should focus on the ethical implications of both emerging AI systems as well as established humanitarian data management practices. Trial registration: Not applicable. Humanitarian data Ethical issues Artificial intelligence (AI) Data processing Ethical risks Scoping review Bioethical principles Resource distribution Figures Figure 1 Figure 2 Background Organizations involved in providing humanitarian assistance work under strenuous circumstances and with limited funding to provide life-saving humanitarian assistance. However, resources for providing this assistance are far from sufficient. In 2022, donor governments provided US $ 27 billion to help 216 million people in 69 countries—a significantly smaller amount than the US $ 51.7 billion required to assist all people in need of humanitarian assistance for that year [ 1 , 2 ]. This considerable shortfall highlights the urgent need to better assess humanitarian needs and to do so at minimal cost. The aim of this review is to map the range of ethical issues that have been raised in the academic literature regarding data processing of people affected by humanitarian crises. Humanitarian organizations rely on processing increasingly large amounts of data to inform their operations, much of which is collected directly from affected populations (e.g., through registrations, household surveys, or cash disbursements). At the same time, the people working for these organizations have themselves often become targets of kidnappings and killings, which has led organizations to increasingly resort to remote methods of managing operations and collecting data from affected people [ 3 , 4 ]. The COVID-19 pandemic has accelerated this trend of the increased use of remote methods. This combination of factors has led to an exponential increase in the amount of personal data that is being distributed, stored, and analyzed in various locations around the world. At the same time, humanitarian organizations are continuously seeking innovations involving information and communication technologies (ICT) in the pursuit of operational gains in effectiveness and efficiency. This practice is expected to accelerate further with the growing availability and sophistication of artificial intelligence (AI) technologies in the health and humanitarian sectors. Definitions A review by Schofield et al. [ 5 ] found that the vast majority of included studies discussing “ethical challenges” in healthcare had failed to include an explicit definition of how that term was understood by the respective authors, leading to potential misunderstandings and ambiguity. This section, therefore, will first provide working definitions for the key terms and concepts discussed in this study. Humanitarian assistance is understood here to refer to coordinated actions that save lives and alleviate suffering of crisis-affected populations [ 6 ]. It also includes “protection”, which “encompasses all activities aimed at obtaining full respect for the rights of the individual in accordance with the letter and the spirit of the relevant bodies of law” [ 7 ]. Humanitarian crises are defined here as a “series of events representing a critical threat to the health, safety, security or wellbeing of a community, usually over a wide area” [ 8 ]. For the purposes of this study, data processing is understood as: “Any operation or set of operations which is performed on data or on sets of data, whether or not by automated means, such as collecting, registering, storing, adapting or altering, cleaning, filing, retrieving, using, disseminating, transferring and retaining or destroying” [ 9 ]. Ethical issues are defined in this study as actions that may not conform to moral standards, particularly those set out by various humanitarian principles [ 10 ] because of the risks they present. Context Organizations turning to new or existing digital tools to collect, store, or analyze data more efficiently may knowingly or inadvertently introduce new ethical issues affecting people who are already vulnerable [ 11 ]. Weighing the responsible use of new technologies in humanitarian crises is fraught with a number of ethical issues [ 12 ] that are increasingly being highlighted in specific circumstances such as refugee registrations [ 13 ], health emergency response [ 14 ], or the use of drones in humanitarian assistance [ 15 ]. In practice, ethical decisions are made—knowingly or unknowingly—on a daily basis about what data to collect, which tools to use, or how and with whom to share this information to avoid adverse consequences [ 16 , 17 ]. In light of such challenges, organizations rarely choose to forego new tools altogether, such as Oxfam’s decision in 2015 to halt the use of biometrics in its programs in order to assess the potential risks [ 18 ]. Rather, some organizations are more likely to invest in new innovations without considering, weighing, or fully grasping the long-term ethical issues [ 19 ]. However, because of these challenges, more guidelines are now being produced for the ethical processing of data for humanitarian assistance purposes, with the goal of minimizing or eliminating risks to vulnerable people. Notable examples include Data Responsibility in Humanitarian Action by the Inter-Agency Standing Committee [ 20 ], the Handbook on Data Protection in Humanitarian Action by the International Committee of the Red Cross [ 21 ], and the Data Responsibility Guidelines by the United Nations Office for the Coordination of Humanitarian Affairs [ 9 ]. Similarly, regulatory environments are changing in many countries (such as the European Union’s General Data Protection Regulation, GDPR), which have moved many humanitarian organizations to change their approaches to data processing in order to improve data privacy. Focusing on the issue of ethical design of new tools, Krishnaraj et al. [ 22 ] have created practical guidelines that aim to mitigate risks as early as possible [ 23 ]. But the speed of technological innovation means that such guidance can quickly become out of date as new data technology tools appear and organizations respond to new circumstances (such as insecure environments or lack of access to populations during the COVID-19 pandemic). Artificial intelligence (AI) systems that use machine learning and other methods for automating data processing may usher in a completely new set of ethical issues that humanitarian organizations will have to confront [ 24 ]. Innovations using AI in the medical and health sectors has been growing significantly for years and is showing important premises, such as in the discovery of new classes of antibiotics [ 25 ]. At the same time, large language models such as ChatGPT [ 26 ] that excel at generating and summarizing human language are generating yet another set of novel ethical issues [ 27 ], including in the health and medical sectors [ 28 , 29 ]. Although a considerable number of studies discuss the ethics of using various technologies in humanitarian assistance, to date, there is little evidence that there has been a comprehensive review of relevant ethical issues in the published literature. Humanitarian and Technical Nomenclatures Conducting this type of review is challenging due to the wide-ranging nature of humanitarian assistance, lack of well-defined nomenclature for data processing technologies and activities, and that relevant research may be published in the intersecting fields of ethics research, design, engineering, health, medicine, geography, development, social science, and technology research, among others. Previous scoping reviews focusing on humanitarian assistance only addressed more limited contexts, such as natural disasters [ 30 ], and displaced populations [ 31 ], or did not include terms to capture more novel humanitarian activities such as responding to large-scale migration or public health emergencies [ 32 ]. Other relevant studies in the past have included a literature review focused on social media and privacy issues (based on literature published between 2013 and 2014 [ 33 ], a scoping review on the types of digital tools used during the 2014–2016 Ebola outbreak in West Africa [ 34 ], a scoping review on the impact of health-related tools in humanitarian crises [ 35 ], and a scoping review on ethical considerations related to the use of drones in humanitarian assistance [ 36 ]. However, we have not found a sufficiently comprehensive sets of keywords that could be used to search databases for any of this study’s three inclusion criteria (people affected by humanitarian crises, processing data for humanitarian assistance, meaningful discussion of ethical issues). Even though the data collection for this study was conducted in 2020, it remains the most recent scoping review on this topic. Ethical Frameworks Another distinct challenge is the lack of established ethical categories or theories used by studies discussing ethics in the humanitarian sector [ 36 ]. First introduced by the International Federation of Red Cross and Red Crescent Societies [ 37 ], the four humanitarian principles (humanity, impartiality, independence, and neutrality) are now widely used among many humanitarian organizations [see, for example, 38], in international law [ 39 ], as well as in ethical codes attempting to guide the actions of the humanitarian sector as a whole [see, e.g., 40, 41]. However, previous studies have shown the difficulty of applying these humanitarian principles in everyday practice [ 42 ], in guiding the use of information technology [ 43 ], or in mapping humanitarian organizations’ ethical obligations [ 44 ]. In particular, the broad humanity “principle” has been argued as being better understood as an absolute moral value rather than an ethical principle [ 10 ]. Developed in parallel, the ethical value categories of autonomy , beneficence , non-maleficence , and justice , are widely used in the fields of bioethics and research ethics, and have been defined and discussed in their application in detail by Beauchamp and Childress [ 45 ]. A definition of each ethical value category is provided in Table 1 . The four categories reflect work largely in the decades following World War II that have aimed to better protect research participants, including the Nuremberg Code [ 46 ], the 1964 Declaration of Helsinki [ 47 ] and the Belmont Report [ 48 ]. Table 1 Definitions of Each Ethical Value Category, Based on Beauchamp and Childress [ 45 ] Principle Definition Respect for autonomy Respecting the decision-making capacities of autonomous persons Beneficence Providing benefits and balancing benefits against risks and costs Non-maleficence Avoiding the causation of harm Justice Distributing benefits, risks, and costs fairly As a result of the challenges in applying humanitarian ethical principles, a growing number of studies use the four ethical value categories as a better operational ethical terms to reference ethical issues inherent to humanitarian practice [ 49 , 50 ]. We also chose to employ these four ethical value categories to group the ethical issues identified in the literature and to better link the nascent field of humanitarian ethics to the larger theoretical and practical advances in the fields of bioethics and research ethics. The aim of this review is to map the range of ethical issues that have been raised in the academic literature regarding processing relevant to people affected by humanitarian crises. This study contributes to the existing academic discussion in three important ways. First, this study presents the first comprehensive review of the ethical considerations in processing data from individuals affected by humanitarian crises, addressing a significant gap in the literature. Second, it addresses the challenges of fragmented terminology by establishing an evidence-based search strategy to cover topics in the intersection of humanitarian assistance, data processing, and ethical implications. Third, the study introduces a clear, transparent framework for defining what constitutes a “humanitarian crisis,” providing a consistent basis for the inclusion or exclusion of different studies, which may help avoid subjective biases in research selection for future studies. Methods Study Protocol We chose to conduct a scoping review as this method is best suited for generating a broad overview of relevant evidence, to examining emerging areas of research, to clarifying key concepts, and to identifying gaps in the literature [ 51 ]. A study protocol was developed prior to data collection and screening using the scoping review method established by Arksey and O’Malley [ 52 ] as further refined by Levac et al. [ 53 ] and follows the framework maintained by the Joanna Briggs Institute [ 54 ]. The protocol was revised based on feedback received from the research team and incorporated the results from a pilot conducted for this study November-December, 2019. It follows the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) reporting guidelines [ 55 ]. The final version of the PRISMA-ScR checklist and the study protocol are available in Appendix A and Appendix B, respectively. Identifying the Research Question The specific research questions of this scoping review were: 1. Which ethical issues have been raised in the literature related to processing data from people affected by humanitarian crises in order to inform humanitarian assistance? 2. To what extent do real-world examples of ethical issues reflect the concerns presented in the literature? 3. Which technologies were the focus of concern over these ethical issues? Eligibility Criteria The following eligibility criteria for the selection of relevant studies were established a priori as per the categories and requirements for scoping review protocols [ 55 ]. Condition/Domain : Ethical issues stemming from the processing of data relating to people affected by a humanitarian crisis with the explicit goal or potential of informing humanitarian assistance. Population : People affected by a humanitarian crisis, including armed conflicts, natural disasters, and large public health emergencies, as well as refugees and transborder migrants fleeing from such a crisis—regardless of their current location. We also included studies that concern humanitarian assistance (including related fields such as disaster response or emergency management) that are global in scope. Studies about natural disasters were only included if the study focused on events in low or lower middle-income countries (defined as countries that ranked low income or low middle income at least once by the World Bank between 2011–2020) [ 56 ]. The Ebola outbreak in West Africa (2014–2016) was included as it was widely considered to be a humanitarian crisis in scope [ 57 ]. We used the Financial Tracking Service by UN OCHA [ 58 ] to judge if an event should be considered a humanitarian crisis (defined as whether a given country was a recipient of humanitarian aid in the same year). Interventions : Data processing relating to people affected by a humanitarian crisis with the explicit goal or potential of informing humanitarian assistance. Excluded were studies that focus on technologies that do not process data on affected people, such as robotics for clearing debris or land mines, algorithmic models for predicting the occurrence or impacts of natural hazards, or tools used for planning humanitarian logistics (e.g., relief/distribution networks, supply chain management, and resource scheduling). Outcomes : Studies that investigate ethical issues stemming from the processing of data (as defined above) were included only if they contained a significant discussion about this subject. During the screening stage, studies were eligible for inclusion if the abstract referenced or mentioned potential ethical issues. During the full text review this was assessed qualitatively by two reviewers. Study Designs : All study designs were eligible for inclusion, including empirical studies, commentaries, and theoretical papers. Excluded were non-peer reviewed studies as well as book reviews. In order to establish a robust foundation for formulating evidence-based recommendations and for feasibility reasons, this research was limited peer-reviewed literature. Context : For feasibility reasons, we restricted the review to studies published after 1 January 2010. Setting : Studies in all countries or territories affected by a humanitarian crisis (or relevant host countries for refugee or cross-border migrant or displaced populations) were included, as defined above. Search Strategy and Information Sources Comprehensive literature searches of electronic databases were conducted on 31 March 2020 for studies published between 2010 and 2019, using Ovid, Ebsco, Web of Science, and Proquest to search 20 databases for relevant studies. Only studies published in English, French, or Spanish were included. As recommended by the scoping review guidelines described above, keywords were selected and piloted in multiple iterations to identify all relevant articles. We had previously identified 34 studies, and these were used as a minimum search target. After an initial search showed that only 13 were included, we repeated the database search over several iterations with additional terms until all 34 studies were reflected in the results. This yielded additional keywords such as “risks” and “challenges” to represent ethical challenges, as well as “innovation” and “experimentation” which are sometimes used to refer to data processing activities. Further, careful searching for terms such as “acute malnutrition” or “forcibly displaced population” were also found to describe specific phenomena in a humanitarian crisis without using terms such as “refugees” or “humanitarian” in the study’s metadata. Likewise, to find all studies that discuss processing data of affected people, we iteratively expanded our search terms to include specific technologies (e.g., biometrics, remote sensing), emerging practices (e.g., remote management, crowdsourcing), or shorthand keywords introduced by researchers (e.g., experimentation, crisis informatics, innovation). A sample of the search strategy for the Ovid databases is displayed in Table 2 . The complete search syntax for each database can be found in Appendix C. Table 2 Search Strategy for Ovid Databases Concept Keyword and syntax Humanitarian assistance 1 humanitarian*.tw. 2 relief work.tw. 3 aid work.tw. 4 (disaster? Adj (relief or response? Or assistance)).tw. 5 emergency relief.tw. 6 ((conflict? Or war?) adj10 (human rights or public health)).tw. 7 (ebola adj6 (west africa or sierra leone or liberia or guinea or 2014 or 2013)).tw. 8 acute malnutrition.tw. 9 (refugee* adj2 (camp* or assistance or population?)).tw. 10 (displace* adj2 (forced or forcibly or population? Or human? Or internal*)).tw. 11 (((population? Or person* or communit*) adj3 affected) adj1 (conflict? Or violence)).tw. 12 or/ 1–11 13 (cris?s or emergenc* or disaster?).tw. 14 humanitarian*.af. 15 13 and 14 16 12 or 15 ICT for data collection 17 ict.tw. 18 technolog*.tw. 19 ((data or information) adj2 (system* or manage* or collection or analys?s or process*)).tw. 20 (blockchain or distributed ledger).tw. 21 (ai or artificial intelligence or machine learning or algorithm*).tw. 22 biometric*.tw. 23 smartphone app*.tw. 24 remote sensing.tw. 25 analytics.tw. 26 digital*.tw. 27 experimentation.tw. 28 automat*.tw. 29 innovation?.tw. 30 remote management.tw. 31 cyber.tw. 32 big data.tw. 33 (sms or text messag* or interactive voice recognition or online survey*).tw. 34 (kobotoolbox or kobo or odk or open data kit).tw. 35 crowdsource*.tw. 36 social media.tw. 37 crisis adj (informatics or data or map*).tw. 38 digiti?ation.tw. 39 datafication.tw. 40 or/ 17–39 Ethical concerns 41 concern?.tw. 42 risk?.tw. 43 challenge?.tw. 44 harm?.tw. 45 privacy.tw. 46 protection?.tw. 47 humanitarian adj (principle? Or standard? Or guideline?).tw. 48 problem?.tw. 49 bias?.tw. 50 ethic*.tw. 51 consequence?.tw. 52 critique?.tw. 53 insecurity.tw. 54 implications.tw. 55 peril?.tw. 56 impact?.tw. 57 or/ 41–56 16 and 40 and 57 Study Selection Study selection and coding were done using the DistillerSR systematic review software [ 59 ]. Using the a priori eligibility criteria, we developed questionnaires for selecting citations during discrete title, abstract, and full text review stages. Two reviewers independently selected studies during each screening stage. Regular meetings to discuss rating discrepancies and to compare working definitions were held during the review of the first 1,000 references in the title screening stage and for the first 100 references during the abstract screening stage. Any conflicts during the title and abstract screening stages were included in the full text review. In the full text screening stage, daily meetings were held during the review of the first 20 references to discuss rating discrepancies and to improve working definitions of terms. Rating discrepancies were resolved by discussion, and in five cases, by using a third adjudicator. Data Collection Process For included studies, we extracted details on study characteristics (year of publication, countries of all authors, author organization types), population characteristics (type of humanitarian crisis), intervention characteristics (purpose of data processing, technologies described), and outcomes (specific ethical issues identified, whether studies used real-world examples to identify issues). Author organization types were coded for all listed affiliations, while author country was extracted only from the first-listed affiliation. For each country, we additionally tabulated the geographic region and income level, using the 2020 World Bank classification scheme [ 56 ]. The data extraction form was created in the DistillerSR software. It was then piloted based on a random sample of 10 included studies and modified based on discussions and feedback from the two reviewers. As per the study protocol, since the number of included citations was greater than 30, data extraction was done by one reviewer and verified by another. The data extraction form included several pre-coded ethical issues, but additional emergent issues could be entered qualitatively in text format. Synthesis We summarized results quantitatively (using frequencies) and qualitatively (using descriptive analytics). We analyzed and coded the ethical issues related to data processing that were entered in text form using SPSS 25. Specific issues described by authors could be assigned to one or more categories of ethical issues. Issue codes were updated iteratively and recursively by creating new codes based on new observations and through constant retrospective reviews of previously collected data. In some cases, rarely-mentioned codes were also merged retrospectively to limit the size of the final list of issues. The ethical issues mentioned in each study were then grouped into the ethical value categories of autonomy, beneficence, non-maleficence, and justice, based on which category was deemed to be the affected most. Results Literature Search The database literature search returned 8,387 citations (see Fig. 1 ). After removing duplicates, 5,999 were included for screening. 3,951 were excluded during the title screening stage and 1,752 during abstract screening. After reviewing full texts of 296 potentially relevant studies, 198 were excluded. As a result, 98 were included in this scoping review (full list of citations listed in Appendix D). Study Characteristics The included 98 studies were published between 2010 and 2019, as shown in Table 3 . The majority (n = 72) were published after 2015, and the most common publication year was 2019 (n = 28). Most were written by authors based in Europe and Central Asia (n = 55) and North America (n = 37), while only a small number of studies included authors from East Asia & Pacific (n = 8), South Asia (n = 4), Sub-Saharan Africa (n = 3), Middle East and North Africa (n = 3), and Latin America and the Caribbean (n = 2), as shown in Fig. 2 . 31 studies included an author from the United States while about one quarter (n = 27) included an author from the United Kingdom. Overall, 92 studies included at least one author from a high-income country while a smaller number included at least one author from an upper middle-income country (n = 7) or lower middle-income country (n = 7). No study included an author from a low-income country. Similarly, no study included an author from China. The vast majority (n = 90) of studies included at least one author from an academic institution, while only 7 studies included at least one author affiliated with a humanitarian organization. Table 3 Study Characteristics (n = 98) Characteristic Count (%) Year of publication 2010 2 (2%) 2012 1 (1%) 2013 5 (5%) 2014 9 (9%) 2015 9 (9%) 2016 20 (20%) 2017 11 (11%) 2018 13 (13%) 2019 28 (29%) Region represented by authors Europe & Central Asia 55 (56%) North America 37 (38%) East Asia & Pacific 8 (8%) South Asia 4 (4%) Sub-Saharan Africa 3 (3%) Middle East & North Africa 3 (3%) Latin America & Caribbean 2 (2%) Country income level based on author location High income 92 (94%) Upper middle-income 7 (7%) Lower middle-income 7 (7%) Low-income 0 (0%) Parent organization type based on author affiliation Academic 90 (92%) For-profit 12 (12%) Non-profit 7 (7%) Humanitarian 7 (7%) Think tank 2 (2%) Government 1 (1%) Type of Humanitarian Crisis Similar numbers of studies focused on or included examples of natural disasters and armed conflict (n = 37 and n = 35, respectively), as shown in Table 4 . Of the 98 studies selected, 31 discussed people displaced by a humanitarian crisis, whereas 19 focused on large public health emergencies. Twenty studies were general in nature and only discussed the fields of humanitarian assistance, emergency management, or disaster response without providing specific examples. Table 4 Types of Humanitarian Crises Discussed (n = 98) Type of humanitarian crisis Count (%) Natural disaster 37 (38%) Armed conflict 35 (36%) People displaced by a humanitarian crisis 31 (32%) Large public health emergency 18 (18%) Not specified 19 (19%) Purpose of Data Processing While most studies reported more than one purpose, the most common data processing purpose was conducting assessments (n = 35), such as needs assessments or damage surveys (see Table 5 ). Twenty-four studies examined different forms of case management (e.g., refugee registrations), while 18 discussed handling of medical or public health data. Twenty-three did not specify any reasons for data processing but instead discussed in theoretical terms the use of information and communication technologies or data processing in humanitarian assistance. Table 5 Data Processing Purposes and Technologies Described by Studies (n = 98) Purposes of data processing Count (%) Assessment (of needs, damage, etc.) 35 (36%) Registration / case management 24 (24%) Forecasting / modeling / early warning 18 (18%) Medical care or public health 18 (18%) Delivery of assistance 12 (12%) Accountability (complaints, feedback collection, etc.) 9 (9%) Cash transfer 8 (8%) Logistics 8 (8%) Search and rescue 8 (8%) Human rights violations 7 (7%) Other 17 (17%) Not specified 23 (23%) Technologies Described The most commonly described technologies used for data processing were social media (discussed by 47 studies), crowdsourcing (n = 44), various forms of mapping and other forms of geographic information systems (GIS; n = 42), whereas nearly one in three studies focused on big data (n = 29) or private messaging services (n = 29). One quarter (n = 25) discussed various forms of tools and technologies related to artificial intelligence (AI), including the use of algorithms and machine learning. The capture of satellite images for humanitarian assistance and the collection of biometrics (typically, fingerprint or iris scans) were discussed by 22 and 21 studies, respectively. Other technologies cited are shown in Table 6 . Five studies did not discuss any specific technologies used for data processing. Table 6 Technologies Described by Studies (n = 98) Specific technologies described Count (%) Social media 47 (48%) Crowdsourcing 44 (45%) Mapping / GIS 42 (43%) Big data 29 (30%) SMS or private messaging software 29 (30%) AI / algorithms / machine learning 25 (26%) Satellite imagery 22 (22%) Biometrics 21 (21%) Information systems 21 (21%) Drones / UAV 18 (18%) Cash distribution 11 (11%) Medical data 10 (10%) Call data records 9 (9%) Data storage 5 (5%) Blockchain / distributed ledger technology 4 (4%) Computer-assisted personal interviewing 3 (3%) Not specified 5 (5%) Ethical Issues Identified As shown in Table 7 , we identified 22 ethical issues in the studies under investigation, which were grouped according to the four previously identified bioethical values categories. Eight issues were attributed to the ethical value category of autonomy, six to beneficence, seven to non-maleficence, and five to justice. On average, studies cited seven different ethical issues each ( M = 6.86, SD = 3.32), ranging from more than two for non-maleficence issues ( M = 2.43, SD = 1.3) to less than one for justice issues ( M = 0.96, SD = 0.76; see Table 8 ). The vast majority of studies mentioned issues related to non-maleficence (n = 91) and beneficence (n = 87). Slightly fewer studies discussed issues concerning justice (n = 71) and autonomy (n = 69). Table 7 Ethical Issues Identified (n = 98) Ethical issues identified Count (%) Autonomy Lack of consent: Data is collected without informed consent 50 (51%) Data agency: People do not have the right to control, access, or delete their data 30 (31%) Lack of respect: People/communities are not treated with respect 29 (30%) Autonomy: Unwillingness to share data does not lead to disadvantages (e.g., exclusion from assistance or protection 24 (24%) Participation: People/communities are not involved in decisions to use of new/experimental technologies for collecting data 16 (16%) Undisclosed use: Data may be used beyond purposes for which they were collected 8 (8%) Lack of group agency: Processed information is not available to affected communities 6 (6%) Any implication related to Autonomy 69 (70%) Beneficence Unreliability: Processed data is inaccurate and does not sufficiently reflect reality to inform assistance 60 (61%) Dependence: Data is processed with the assistance of a political, economic, or military entity 47 (48%) Lack of action: Processed data is not utilized to inform assistance to the affected person/community 38 (39%) Non-neutrality: Data is processed in a way that benefits or appears to benefit one side of the conflict over the other 28 (29%) Ineffective or inefficient: Not producing expected result, unmet expectations 4 (4%) Any implication related to Beneficence 87 (89%) Non-maleficence Privacy: Personal/sensitive data is shared with third parties 72 (73%) Harm: People suffer physical or psychological harm as a result of data processing 66 (67%) Data security: Personal/sensitive data is not protected against malicious actors 41 (42%) Power imbalance: Data processing reinforces or worsens a lack of power of affected people 33 (34%) Excess: More data was collected than necessary 17 (17%) Redress/rectification: People do not have the ability to correct wrong information about them or receive compensation 9 (9%) Any implication related to Non-maleficence 91 (93%) Justice Bias: Data is processed in a way that may (dis)advantage some people disproportionate to their humanitarian needs 61 (62%) Lack of accountability: Endangering (or not protecting) rights; absolving responsibility 16 (16%) Unequal access to technology / exclusion from data collection 14 (14%) Unfair distribution of risks and benefits 3 (3%) Any implication related to Justice 71 (72%) [insert Table 7 here] Table 8 Number of Ethical Issues Cited by Ethical Value Category (n = 98) Bioethical value category mean, SD (min to max) All ethical value categories 6.97, 3.4 (1 to 15) Autonomy 1.73, 1.59 (0 to 7) Beneficence 1.82, 1.16 (0 to 5) Non-maleficence 2.44, 1.3 (0 to 6) Justice 0.98, 0.77 (0 to 3) The most frequently cited ethical issue categorized under the value category of autonomy was data being collected without sufficient informed consent (n = 52). For example, Shoemaker et. al. [ 60 ] found through interviews with refugees are frequently being asked by humanitarian organizations to provide personal information that the respondents considered intrusive, without being offered a justification on why this was relevant. Within the value category of beneficence, the ethical issue most frequently mentioned by studies was processed data being inaccurate and not sufficiently reflecting reality to inform assistance (n = 61). This is illustrated by Paul and Sosale [ 61 ], who cite the challenges of using social media as a basis to inform humanitarian assistance. In an example the authors cite, the same information was re-posted multiple times by well-meaning users, making it difficult for emergency responders after a severe flooding event to identify new information that might require a team to be dispatched. Falling under the value category of non-maleficence, the most-cited ethical issue (n = 74) were privacy concerns in cases where personal or sensitive data may be shared with third parties. For example, Hayes and Kelly [ 62 ] discuss how personal requests for help that are aggregated by a crowdsourcing platform such as Ushahidi can make personal information publicly available, including to bad actors trying to exploit vulnerable people. The most frequently mentioned ethical issue categorized under the justice value category was biased data processing leading to (dis)advantaging people disproportionate to their humanitarian needs (n = 63). This issue, which often relates to different forms of sampling problems that could endanger the impartial distribution of aid, has become more pressing as more organizations turn to “big data” solutions for informing humanitarian assistance without properly understanding their limitations [ 63 ]. Information Sources for Ethical Issues As shown in Table 9 , slightly over half of studies (n = 52) cited at least one real-world example of an ethical issue, usually based on anecdotal information found in news reports or other published literature [see, e.g., 64, 65]. Fully 29 studies included ethical issues that were raised by interviews or other kinds of consultations with experts. Examples here include Shoemaker et. al. [ 60 ], who conducted qualitative interviews with 198 refugees in Lebanon, Jordan, and Uganda, as well as Vannini et al. [ 66 ], who interviewed nine representatives from organizations assisting transborder migrants in the United States. Four studies included a systematic review of the literature [ 33 – 35 , 67 ]. Table 9 Information Sources of Ethical Issues (n = 98) Sources of ethical issues Count (%) Specific instances of ethical issue (rooted in real-life experience) 52 (53%) Ethical concerns raised in interviews/expert consultations 29 (30%) Systematic review of the literature 4 (4%) Key Results for Studies Discussing AI Of the 25 studies that discuss the use of AI, all were published since 2014, with about half (n = 13) published after 2017 (see Appendix E for all figures related to the AI-related studies). The most common type of humanitarian crisis discussed in the 25 studies was natural disaster (n = 9), followed by armed conflict and large public health emergencies (n = 7, respectively). The most common purposes for data processing were related to assessments (n = 9) as well as handling medical or public health data (n = 7). The majority of the 25 studies related to AI also discussed big data (n = 15), social media (n = 15), and GIS (n = 13). The vast majority mentioned ethical implications related to privacy (n = 23) and the risk of physical or psychological harm (n = 22). Eighteen studies related to data being processed in a way that may result in aid being distributed disproportionately with regard to people’s actual needs. Discussion The aim of this review was to map the range of ethical issues that have been raised in the academic literature regarding data processing of people affected by humanitarian crises. This review identified 22 such ethical issues. Issues related to the value category of non-maleficence were brought up by the vast majority of studies (n = 93), which dovetails with a strong trend in the recent literature focusing on the imperative of “do no harm” in humanitarian assistance [ 68 – 70 ]. The risk of increasing harm (whether physical or psychological) as a result of data processing was mentioned by a high number of studies (n = 68). Privacy concerns were cited by 74 studies—far more commonly than all other issues—reflecting an increased awareness of this issue over the last years across organizations and the media. Among studies discussing AI as a data processing technology, privacy concerns were even more prevalent, with 23 out of 25 (92%) mentioning this issue. This points to a significant worry across the humanitarian sector about the many ways in which personal data from affected people is being processed in a manner that may endanger their right to privacy, which is enshrined in the 1948 Universal Declaration of Human Rights [ 71 ]. Studies discuss a wide gamut of how personal privacy can be violated, including accidental or intentional sharing with third parties beyond what the affected person had agreed to during personal interviews, or if at all. Even in cases where informed consent was given, interviewees in vulnerable situations—or who lack understanding of sophisticated data management, access and processing—may not understand all the potential ways in how their personal information may be used, stored and accessed. Collecting and processing personal data from social media, by unmanned aerial vehicles (UAV), or public records (often under the “big data” category) that lack explicit consent are particularly problematic. Although the protection of privacy can be understood an essential right to safeguard human dignity [ 72 ], more studies and initiatives in humanitarian assistance need to resolve the apparent conflict between the duty to protect privacy and the urgent duty to assist and protect those in danger [ 62 ]. Many studies pointed out that organizations frequently collect much more data than they need (n = 17) or are able to absorb (n = 40). We consider the former as a potential for harm, as any excessive information increases the risks associated with data leaks and privacy violations. Data collected but that is not used can be regarded as related to the ethical value of beneficence, which implies that all information collected should have a concrete purpose related to informing humanitarian assistance. But even for data that were used for the intended purpose, a majority (n = 61) of studies discussed that it may be too unreliable or inaccurate to adequately inform assistance programming. A strong theme emerged regarding insufficient consent mechanisms, which strongly relates to the ethical value category of autonomy. About half the studies (n = 52) mentioned that informed consent was either not provided by the affected population or was given without a full picture of how data would be processed or used. Eight studies cited that data might be used for reasons other than the original purpose for which consent may have been obtained. Related to the ethical value category of autonomy, about one in four papers (n = 26) mentioned that a refusal to provide information could lead to being excluded from receiving assistance. This issue is illustrated by Shoemaker et al. [ 60 ] who documented how refugees felt that they lacked a choice on whether or not to provide personal information to UNHCR as their ability to access assistance depended on it. Detailed guidance has been created by the International Rescue Committee on how to obtain proper consent, whereas the International Committee of the Red Cross has published the legal basis for situations when data processing is permissible—even when consent cannot be assumed or obtained [ 21 , 73 ]. However, more work is clearly needed to train humanitarian professionals in these practices, and to monitor for better compliance with best consent practices as well as other minimal ethical guidelines. Existing guidance also needs to be updated to ensure the protection of private, personal and demographically identifiable information that extends to population groups rather than individuals [ 74 ]. Directly related to the value category of justice, a majority (n = 63) of studies were concerned about data being processed in a way that may result in aid being distributed disproportionately with regard to people’s actual needs. This finding directly mirrors the importance of the humanitarian impartiality value category which refers to providing assistance solely based on need and regardless of personal preferences or discriminatory factors [ 37 ]. A cross-cutting issue was the potential of data processing to exacerbate power imbalances (mentioned by 34 studies), often due to an exclusion from data collection, given the unequal access to certain technologies (n = 14). In many cases, data processing was found to diminish the perceived neutrality of humanitarian organizations (n = 29) as data could be processed in a way that might benefit one side of the conflict over the other. Concerningly, about half (n = 48) of the studies found that humanitarian data processing might be overly dependent on potentially biased external entities (such as commercial entities, militaries, or foreign governments). This could be increasingly problematic for humanitarian organizations for multiple operational and ethical reasons, but particularly in conflict environments where the perception of independence is widely considered to be an essential humanitarian value category. Another theme identified across many studies was that data processing did not follow the principles of Accountability to Affected People (AAP) [ 75 , 76 ], which manifested in various ways across several of the four bioethical value categories. For the value category of autonomy, 18 studies remarked that affected communities were not involved in decisions on whether to use experimental technologies, whereas a smaller number (n = 8) commented that processed information was not being made available to communities to allow for better group agency. Related to the values category of non-maleficence, nine papers discussed people’s inability to rectify inaccurate information about them or receive any form of compensation. Finally, related to the value category of justice, one in six (n = 16) of studies found that data processing lacked accountability in terms of humanitarian organizations’ obligation to protect rights—or even pointed to ways that they may be violating these rights themselves. The Signal Code [ 77 ], first published by the Harvard Humanitarian Initiative in 2017, considers data agency and redress/rectification as crucial rights and proposes specific actions to safeguard them in practice. We propose extending this list to always include affected communities when sharing collected data and involving them in decisions over experimental technologies. About half of studies (n = 52) cited ethical issues that were rooted in real-life experiences whereas almost one third (n = 29) contained issues based on qualitative interviews or expert consultations. This signals that ethical issues have moved from theoretical concerns to actual incidents. However, it also reflects the large and diverse array of ethical issues that are emerging in connection with data processing in humanitarian crises which may first manifest as theoretical concerns before being validated as potentially negative consequences that can and do occur in real life. The results from this study show a wide array of ethical issues that should be addressed when processing data in a humanitarian context. However, to our knowledge, to date no humanitarian data protection guidance is sufficiently comprehensive to provide practical guidance for all concerns identified in the literature. More work may therefore be needed to expand existing guidelines and to ensure that future updates are also informed by systematic reviews of newly published studies. Moreover, there is a need to improve humanitarian organizations’ accountability to affected populations, including the need to proactively prevent harm, and monitor the potential for causing harm and to limit risks. Finally, organizations involved in providing humanitarian assistance should review internal processes for training staff and create methods for verification that ensure appropriate minimal ethical standards are being met. Geographically, publications were disproportionately from authors in high-income countries, primarily in Europe and North America, demonstrating a high level of interest in countries that have been the traditional source of most humanitarian funding but also of most technological innovation. Conversely, the small number of authors from lower-middle income countries and the complete absence of any authors from low-income countries highlights the lack of published perspectives from countries most affected by humanitarian crises. The lack of any study with an author from China may reflect that the large body of disaster related studies from Chinese authors published in English primarily discuss the response to domestic rather than foreign crises, or that ethical issues explored in this review may be more explored in Chinese language publications. People living in affected countries make up the vast majority of humanitarian organizations’ staff, which could be a potential boon to a more diverse authorship on this subject. However, given the very small number of studies with authors from a humanitarian organization, more efforts need to be made by publishers to invite and support submissions from humanitarian professionals. From 2010 to 2019, the number of peer-reviewed articles and conference proceedings about the ethical issues related to processing data of people affected by humanitarian crises has grown significantly, particularly after 2015. Published research shows widespread concerns that data processing in humanitarian assistance, without due ethical attention to known harms or potential risks, can cause additional harm, may not provide direct benefit, can show a lack of respect for affected populations’ autonomy, and can lead to the unfair distribution of resources, among other concerns. We found that studies containing ethical discussions are often skewed towards investigating natural disaster contexts, as well as the use of technologies that allow the involvement of non-traditional actors, especially by gathering information from social media or crowdsourcing platforms. More research into ethical issues that arise in conflict settings is needed to better investigate the heightened security risks to vulnerable people in the large number of humanitarian crises associated with conflict, war, and social instability. As some studies in this review have pointed out, data may be collected in many humanitarian crises without due attention to data privacy and the security of the person providing the data, leaving them subject to malfeasant actors. Further, data can be collected or processed in a way that is inaccurate or out-of-date for the purpose for which it is being used, a risk that increases with methods such as crowdsourcing or analyzing social media posts. Both potential outcomes may pose added safety and security risks for people already vulnerable in a humanitarian crisis. They may also potentially render inaccurate needs assessments to humanitarian actors that may mean material humanitarian assistance delivered is inappropriate to actual needs, thus further increasing the vulnerability of people already at risk. The ethical issues identified in this review should be used to inform the development of ethical codes of conduct (whether voluntary or mandated by organizations). Further, companies and institutions behind the various technologies—as well as the humanitarian organizations that use them to process data as part of their work—should investigate to what extent these ethical issues are being addressed, and where more needs to be done. However, existing humanitarian data protection guidance and mechanisms are not sufficient to address all concerns identified in the literature and in this study. Likewise, training and accountability mechanisms to monitor the actual harm or potential for causing harm and to limit risks, are insufficient. These guidelines and mechanisms will need to be reviewed, expanded and informed by regular reviews that keep pace with technological change and changes in practice. Further research, especially using empirical methods, is necessary to better identify and understand the type and prevalence of ethical issues in the field. Finally, more investigations are needed into the appropriate and inappropriate use of commonplace humanitarian tools and data management processes, such as CAPI, spreadsheets, filesharing, or use of online databases. At the same time, case studies of early adoptions of AI should address which ethical considerations were given when using tools that may involve data processing using multiple services and companies globally, in order to inform local decisions. Such research is urgently needed to create better guidance, training, and auditing methods to support humanitarian organizations to use data processing technologies as ethically as possible. The number of studies discussing natural disasters (n = 37) was about the same as the number discussing armed conflict (n = 36), even though by the end of 2020, 87% of displacement was caused by conflict [ 78 ]. This disproportionate focus may be due to disasters generating a higher level of media attention, as well as interest among technology enthusiasts, volunteers, and private companies—a trend identified by several studies [ 79 – 82 ]. Likewise, empirical research in conflict settings is far more difficult given the inherent security risks, which in turn limits the development of theories and academic discourse that rely on data from the field. More research will be needed in the future focusing on ethical issues that are unique to conflict settings, as data processing without appropriate consideration of ethical issues in these settings arguably has the potential to cause far greater harm. Our results show a significant focus on both internally and externally displaced populations, particularly those trying to reach Europe or the United States. Likewise, the Ebola virus disease outbreak in West Africa in 2014–2016 was the focus of major international containment response because of the perceived threat of pandemic spread beyond the region, and the subject of a large number of well-funded studies. The increased focus on assistance to displaced populations could be due to the intense media coverage, whereas technological experimentation during the Ebola response was taken to new levels in areas such as processing of call data records without explicit consent [ 83 ]. A significant number of studies discussed ethical issues without going into detail about the particular context: 20 studies discussed humanitarian assistance or crisis response in general, while 23 did not specify a data processing purpose. Given the lack of widely shared understandings of what constitutes terms such as “humanitarian community” or “information and communication technologies”, we recommend that even theoretical papers provide sufficient definitions and examples. We found that studies most commonly discussed activities involving the initial collection of data from affected populations, including assessments, registrations, and health interventions. To some extent, this reflects that a large number of studies investigated the use of crowdsourcing and social media to gain an understanding of a particular humanitarian crisis (see below). It may also be a reflection of the increasing emphasis that humanitarian organizations and their donors have been placing in recent years on establishing an “evidence base” before rolling out assistance programs [ 84 , 85 ]. More research is needed to investigate the link between the potential increase of ethical risk and the push for collecting more needs assessment data. Studies discussing social media (n = 48), crowdsourcing (n = 45), and mapping (n = 43) dominated, often due to the perceived lack of good ground-validated data in humanitarian assistance. There were many use cases of social media, but the most-discussed application was mining public Twitter posts for clues on potential population needs. We also found that many studies focus on the potential use of other “new” technologies, especially if they can be used remotely to assess needs (e.g., satellite imagery, unmanned aerial vehicles, call data records). Crowdsourcing, a method of obtaining information from the general public [ 86 ], was discussed by almost half the studies. Many studies traced their enthusiasm for—or criticism of—crowdsourcing to the creation of the Ushahidi platform (mentioned by 25 studies) in 2007. Similarly, the emergence of digital platform based volunteer networks since the 2010 Haiti earthquake [ 87 , 88 ] can partially explain the large number of studies referencing these tools. Surprisingly, only three studies mentioned computer-assisted personal interviewing (CAPI) tools such as KoboToolbox which has been adopted by a broad range of international and national humanitarian agencies as the tool of choice for needs assessments [ 89 , 90 ]. Similarly, use of spreadsheets were only mentioned by one study as a cause for ethical concern, despite being the main data storage and sharing mechanism of choice for many humanitarian organizations [ 91 ]. Such low-tech data processing means are addressed in recent guidelines, for example, giving guidance on how to remove sensitive data before sharing Excel files with others [ 9 ]. However, more research is needed on current practices and ethical risks associated with these commonly used technologies. The ethical issues associated with biometrics were discussed by a significant number of studies, particularly for the registration of refugees and other migrants by organizations such as the UN High Commissioner for Refugees, UNHCR [see, for example, 13]. In 2015, such concerns led Oxfam, one of the largest international humanitarian NGOs, to put a moratorium on its use of biometrics in order to assess potential risks [ 18 ]. In 2021, this in turn resulted in the creation of a policy intended to ensure that the technology is used ethically within Oxfam’s operations [ 92 ]. Finally, ethics related to AI and similar technologies were discussed significantly more frequently after 2017. This seems to correlate with the growing presence and desire to analyze “big data” resources, in order to learn more about the needs and sentiments of the affected population. For example, big data and medical data were mentioned twice as frequently by studies that discussed AI; call data records were cited 78% more often. More theoretical and empirical research is needed to understand the potential issues that come with applying these rapidly evolving technologies to humanitarian assistance. Limitations of the Scoping Review This study is limited to literature published since 2010 and before January 2020, and it excludes work from non-peer-reviewed sources. As mentioned above, identifying all relevant studies was a significant challenge due to the lack of a shared nomenclature across disciplines for humanitarian assistance, ethical issues, and data processing. As a result, potentially relevant articles that met the inclusion criteria may have been missed. Nonetheless, we believe that our search strategy represents the most comprehensive and inclusive set of keywords to capture studies in the diverse field of humanitarian assistance to date. As suggested by the Arksey and O’Malley framework, a consultation exercise with humanitarian and ethics experts will be organized to present our results, aid knowledge translation, ensure that the results from this study are relevant, and frame a future research agenda. The results of this consultation will be published separately. Conclusions This extensive review of the literature highlights a growing concern over ethical challenges in data processing within humanitarian contexts, including those related to the increasing use of AI. Our findings underscore significant ethical risks associated with data processing in these settings, including potential harm, lack of direct benefits, infringement on populations’ autonomy, and the unfair allocation of resources. Notably, a quarter of the studies reviewed address AI technologies, pointing to privacy concerns as the most frequent ethical issue, especially regarding the inadvertent sharing of sensitive data with third parties. The underrepresentation of perspectives from low and middle-income countries in the academic discourse further exacerbates these challenges, highlighting the urgent need for more diverse and inclusive perspectives. Additional research, especially using empirical methods, is necessary to better identify and understand the type and prevalence of ethical issues in the field. While natural disasters predominate the literature, more studies are needed investigate the unique ethical issues that arise in conflict settings to better address the heightened security risks to vulnerable people in war. Existing humanitarian data protection guidance as well as training and accountability methods for monitoring potential harm and to limit risks are insufficient to address all concerns identified in the literature and in this study. These guidelines and mechanisms will need to be reviewed, expanded, and informed by regular reviews that keep pace with technological change and changes in practice. Likewise, companies and institutions behind the various technologies—as well as the humanitarian organizations that use them to process data as part of their work—should investigate to what extent the ethical issues identified in this study are being addressed, and where more needs to be done. Finally, investigations are urgently needed into early adoptions of AI tools in humanitarian contexts, including the rapid spread of large language models such as ChatGPT, to ensure these technologies are harnessed with utmost ethical rigor, safeguarding the dignity and rights of those in crisis while enhancing the efficacy and fairness of humanitarian responses. Abbreviations Acronym Definition Page AAP Accountability to Affected People 23 CAPI Computer-assisted personal interviewing 26 COVID-19 Coronavirus disease (2019) 4 GDPR General Data Protection Regulation 5 GIS Geographic information system 16 ICT Information and communication technology 4 OCHA United Nations Office for the Coordination of Humanitarian Affairs 9 PRISMA-ScR followed Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews 2 SMS Short messaging service 17 UAV Unmanned aerial vehicles 17 UNHCR United Nations High Commissioner for Refugees 23 Declarations Ethics Approval and Consent to Participate Not applicable Consent for Publication Not applicable Availability of Data and Materials All data generated or analyzed during this study are included in this published article and its supplementary information files. Competing Interests The authors declare that they have no competing interests. Funding Funding from Grand Challenges Canada provided support for the second screening reviewer’s costs during the scoping review. The funder had no involvement in the study’s design, data collection, or analysis. Authors' Contributions TK led the design of the study, conducted the data collection and analysis, and drafted the initial manuscript. JO provided overall supervision and mentorship to ensure the quality and integrity of the research. LA, PV, and JO provided guidance on the scoping review methodology and assessed the study protocol critically. JO and PV contributed to the definitions of ethical value categories and related concepts. AA contributed to the conceptualization of the study in relation to artificial intelligence and helped refine the research questions. All authors read and approved the final manuscript. Acknowledgements Not applicable Authors' Information (optional) References OCHA. Global Humanitarian Overview 2023 [Internet]. Office for the Coordination of Humanitarian Affairs; 2022. Available from: https://reliefweb.int/report/world/global-humanitarian-overview-2023-enaresfr. OCHA. Financial tracking service. 2023. https://fts.unocha.org. Accessed 2 Jul 2023. Humanitarian Outcomes. Aid worker security database. 2021. https://aidworkersecurity.org/incidents/report. Accessed 1 Sep 2021. Donini A, Maxwell D. From face-to-face to face-to-screen: remote management, effectiveness and accountability of humanitarian action in insecure environments. Int Rev Red Cross. 2013;95:383-413. Schofield G, Dittborn M, Selman LE, Huxtable R. Defining ethical challenge(s) in healthcare research: a rapid review. BMC Med Ethics. 2021;22:135. UN OCHA. Glossary of humanitarian terms. Geneva/New York: OCHA; 2003. IASC. Protection of internally displaced persons. Geneva: IASC; 1999. WHO. Risk reduction and emergency preparedness. Geneva: World Health Organization; 2007. OCHA. Data responsibility guidelines. Geneva: UN Office for the Coordination of Humanitarian Affairs; 2019. Slim H. Humanitarian ethics: a guide to the morality of aid in war and disaster. New York, NY: Oxford University Press; 2015. International Committee of the Red Cross and Privacy International. The Humanitarian Metadata Problem: “Doing No Harm” in the Digital Era [Internet]. Geneva; 2018. Available from: https://privacyinternational.org/sites/default/files/2018-12/The Humanitarian Metadata Problem - Doing No Harm in the Digital Era.pdf. Sandvik KB, Gabrielsen Jumbert M, Karlsrud J, Kaufmann M. Humanitarian technology: a critical research agenda. Int Rev Red Cross. 2014;96:219-42. Jacobsen KL. Experimentation in humanitarian locations: UNHCR and biometric registration of Afghan refugees. Secur Dialogue. 2015;46:144-64. Perakslis ED. Using digital health to enable ethical health research in conflict and other humanitarian settings. Confl Health. 2018;12:23. van Wynsberghe A, Comes T. Drones in humanitarian contexts, robot ethics, and the human–robot interaction. Ethics Inf Technol. 2019;22:43-53. UN Global Pulse. Big Data for Development and Humanitarian Action: Towards Responsible Governance: Global Pulse Privacy Advisory Group Meetings [Internet]. New York, NY; 2016. Available from: http://unglobalpulse.org/sites/default/files/Big_Data_for_Development_and_Humanitarian_Action_Report_Final_0.pdf. Raymond NA, Al Achkar Z, Verhulst S, Berens J. Building data responsibility into humanitarian action. OCHA Policy Stud Ser [Internet]. 2016;18. Available from: https://docs.unocha.org/sites/dms/Documents/TB18_Data Responsibility_Online.pdf. The Engine Room and Oxfam. Biometrics in the Humanitarian Sector [Internet]. 2018. Available from: https://www.theengineroom.org/wp-content/uploads/2018/03/Engine-Room-Oxfam-Biometrics-Review.pdf. Parker B. New UN deal with data mining firm Palantir raises protection concerns. 2019. https://www.irinnews.org/news/2019/02/05/un-palantir-deal-data-mining-protection-concerns-wfp. Accessed 25 Aug 2021. IASC. Operational guidance: data responsibility in humanitarian action. Geneva: Inter-Agency Standing Committee; 2021. International Committee of the Red Cross. Handbook on Data Protection in Humanitarian Action (2nd Edition) [Internet]. 2020. Available from: https://www.icrc.org/en/data-protection-humanitarian-action-handbook. Krishnaraj G, Hunt M, Schwartz L, O’Mathúna D, Pringle J, Okhowat A, et al. Ethics for humanitarian innovation: background paper. London: Elrha; 2021. Elhra, Humanitarian Health Ethics. Ethics for humanitarian innovation: Toolkit [Internet]. London; 2021. Available from: https://higuide.elrha.org/wp-content/uploads/2021/09/Elrha-Ethics_for_Humanitarian_Innovation-Toolkit-210913.pdf. Tegmark M. Life 3.0 : being human in the age of artificial intelligence. New York, NY: Alfred A. Knopf; 2017. Wong F, Zheng EJ, Valeri JA, Donghia NM, Anahtar MN, Omori S, et al. Discovery of a structural class of antibiotics with explainable deep learning. Nature. 2023. doi: 10.1038/s41586-023-06887-8. OpenAI. ChatGPT [Internet]. 2024. Available from: https://chat.openai.com/. Stahl BC, Eke D. The ethics of ChatGPT – Exploring the ethical issues of an emerging technology. Int J Inf Manag. 2024;74:102700. Dave T, Athaluri SA, Singh S. ChatGPT in medicine: an overview of its applications, advantages, limitations, future prospects, and ethical considerations. Front Artif Intell. 2023;6:1169595. Shen Y, Heacock L, Elias J, Hentel KD, Reig B, Shih G, et al. ChatGPT and other large language models are double-edged swords. Radiology. 2023;307:e230163. Tansey CM, Pringle J, Davé A, Boulanger R, Hunt M. Earthquakes to floods: a scoping review of health-related disaster research in low- and middle-income countries. PLoS Curr. 2018;10. doi: 10.1371/currents.dis.57d98a902a326361d88d54521e68b016. Makhoul J, Chehab RF, Shaito Z, Sibai AM. A scoping review of reporting ‘Ethical Research Practices’ in research conducted among refugees and war-affected populations in the Arab world. BMC Medical Ethics. 2018;19:36. Pal NE, Eckenwiler L, Hyppolite SR, Pringle J, Chung R, Hunt M. Ethical considerations for closing humanitarian projects: a scoping review. J Int Humanit Action. 2019;4:17. Watson H, Rodrigues R. Bringing privacy into the fold: considerations for the use of social media in crisis management. J Contingencies Crisis Manag. 2017;26:89-98. Bempong NE, Castañeda RRD, Schütte S, Bolon I, Keiser O, Escher G, et al. Precision global health - the case of ebola: a scoping review. J Glob Health. 2019;9:010404. Mesmar S, Talhouk R, Akik C, Olivier P, Elhajj IH, Elbassuoni S, et al. The impact of digital technology on health of populations affected by humanitarian crises: recent innovations and current gaps. J Public Health Policy. 2016;37:167-200. Wang N, Christen M, Hunt M. Ethical considerations associated with “humanitarian drones”: a scoping literature review. Sci Eng Ethics. 2021;27:51. Pictet J. The fundamental principles of the red cross: commentary. Geneva: ICRC; 1979. MSF. Independent, impartial, neutral [Internet]. [cited 2021 Jan 4]. Available from: https://www.doctorswithoutborders.org/who-we-are/principles/independence. OCHA. Humanitarian principles. New York: UN Office for the Coordination of Humanitarian Affairs; 2012. IFRC. The code of conduct for the international red cross and red crescent movement and non governmental organisations (NGOs) in disaster relief. Geneva: IFRC; 1994. Sphere Association. The sphere handbook: humanitarian charter and minimum standards in humanitarian response. Geneva: Practical Action Publishing; 2018. Hilhorst D, Schmiemann N. Humanitarian principles and organisational culture: everyday practice in Meédecins Sans Frontié res-Holland. Dev Pract. 2002;12:490-500. Raymond N, Card B. Applying humanitarian principles to current uses of information communication technologies: gaps in doctrine and challenges to practice. Cambridge, MA: Harvard Humanitarian Initiative; 2015. Broussard G, Rubenstein LS, Robinson C, Maziak W, Gilbert SZ, DeCamp M. Challenges to ethical obligations and humanitarian principles in conflict settings: a systematic review. J Int Humanit Action. 2019;4:15. Beauchamp T, Childress J. Principles of biomedical ethics. New York, NY: Oxford University Press; 2019. Trials of war criminals before the Nuernberg military tribunals under Control Council law no. 10 (Volume 2) [Internet]. Washington, D.C.: U.S. Government Printing Office; 1949. Available from: https://collections.wip.nlm.nih.gov/catalog/nlm:nlmuid-01130400RX2-mvpart. World Medical Association. World medical association declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013;310:2191-4. National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. Belmont report: ethical principles and guidelines for the protection of human subjects of research. Washingtion, D.C.: Department of Health, Education, and Welfare; 1978. Pham P, Vinck P. Technology fusion and their implications for conflict early warning systems, public health, and human rights. Health Hum Rights. 2012;14:106-17. Cawthorne D, Wynsberghe ARV. From HealthDrone to FrugalDrone: value-sensitive design of a blood sample transportation drone. In: 2019 IEEE international symposium on technology and society (ISTAS). Medford, MA, USA: IEEE; 2019. p. 1-7. Peters MDJ, Godfrey CM, Khalil H, McInerney P, Parker D, Soares CB. Guidance for conducting systematic scoping reviews. Int J Evid Based Healthc. 2015;13:141-6. Arksey H, O'Malley L. Scoping studies: towards a methodological framework. Int J Soc Res Methodol. 2005;8:19-32. Levac D, Colquhoun H, O'Brien KK. Scoping studies: advancing the methodology. Implement Sci. 2010;5:69. Peters MDJ, Godfrey C, Mcinerney P, Baldini Soares C, Khalil H, Parker D. Chapter 11: Scoping reviews. In: Aromataris E, Munn Z, editors. Joana Briggs Inst Rev Man [Internet]. 2017. p. 6–24. Available from: https://reviewersmanual.joannabriggs.org/ Tricco AC, Lillie E, Zarin W, O'Brien KK, Colquhoun H, Levac D, et al. PRISMA extension for scoping reviews (PRISMA-ScR): checklist and explanation. Ann Intern Med. 2018;169:467-73. World Bank. World bank country and lending groups. Washingtion, D.C.: The World Bank; 2020. UN News Centre. Interview: senior UN relief official warns no room for complacency until Ebola is eliminated. 2015. https://ebolaresponse.un.org/interview-senior-un-relief-official-warns-no-room-complacency-until-ebola-eliminated. Accessed 1 Jun 2021. OCHA. Financial Tracking Service [Internet]. 2023 [cited 2023 Jul 2]. Available from: https://fts.unocha.org. Evidence Partners. DistillerSR [Internet]. Ottawa, Canada; 2022. Available from: https://www.evidencepartners.com. Shoemaker E, Kristinsdottir GS, Ahuja T, Baslan D, Pon B, Currion P, et al. Identity at the margins: examining refugee experiences with digital identity systems in Lebanon, Jordan, and Uganda. In: Proceedings of the 2nd ACM SIGCAS conference on computing and sustainable societies. New York, NY, USA: ACM; 2019. p. 206-17. Paul S, Sosale S. Witnessing a disaster: public use of digital technologies in the 2015 South Indian floods. Digit J. 2020;8:15-31. Hayes P, Kelly S. Distributed morality, privacy, and social media in natural disaster response. Technol Soc. 2018;54:155-67. Latif S, Qayyum A, Usama M, Qadir J, Zwitter A, Shahzad M. Caveat emptor: the risks of using big data for human development. IEEE Technol Soc Mag. 2019;38:82-90. Comes T, Meesters K, Torjesen S. Making sense of crises: the implications of information asymmetries for resilience and social justice in disaster-ridden communities. Sustain Resilient Infrastruct. 2019;4:124-36. Haworth BT, Bruce E, Whittaker J, Read R. The good, the bad, and the uncertain: contributions of volunteered geographic information to community disaster resilience. Front Earth Sci. 2018;6:183. Vannini S, Gomez R, Lopez D, Mora S, Morrison JC, Tanner J, et al. Humanitarian organizations' information practices: procedures and privacy concerns for serving the undocumented. E J Info Sys Dev Countries. 2019;86:e12109. Cinnamon J, Jones SK, Adger WN. Evidence and future potential of mobile phone data for disease disaster management. Geoforum. 2016;75:253-64. Anderson M. Do no harm: how aid can support peace – Or war. Boulder, CO: Lynne Rienner Publishers; 1999. WHS. The Grand Bargain – a shared commitment to better serve people in need. Istanbul: World Humanitarian Summit; 2016. Sandvik K, Raymond N. Beyond the protective effect: towards a theory of harm for information communication technologies in mass atrocity response. Genocide Stud Prev. 2017;11:9-24. UN General Assembly. Universal Declaration of Human Rights [Internet]. Gen. Assem. Resolut. 1948 [cited 2022 Feb 14]. Available from: https://www.un.org/en/ga/search/view_doc.asp?symbol=A/RES/217(III). Floridi L. On human dignity as a foundation for the right to privacy. Philos Technol. 2016;29:307-12. International Rescue Committee. Obtaining Meaningful Informed Consent [Internet]. New York; 2018. Available from: https://www.principlesinpractice.info/help-library/irc-research-toolkit-obtaining-meaningful-informed-consent. Raymond NA. Beyond “do no harm” and individual consent: reckoning with the emerging ethical challenges of civil society’s use of data. In: Taylor L, Floridi L, VanDerSloot B, editors. Group privacy: new challenges of data technologies. Cham: Springer International Publishing; 2016. p. 67-82. International Committee of the Red Cross. Accountability to Affected People Institutional Framework [Internet]. 2019. Available from: https://www.icrc.org/en/publication/accountability-affected-people-institutional-framework. Global Health Cluster. Operational Guidance on Accountability to Affected Populations (AAP) [Internet]. 2017. Available from: https://www.humanitarianresponse.info/system/files/documents/files/cluster_coordination_reference_module_2015_final.pdf. Greenwood F, Howarth C, Poole D, Raymond N, Scarnecchia D. The signal code: a human rights approach to information during crisis. Cambridge, MA: Harvard Humanitarian Initiative; 2017. IDMC. 2021 global report on internal displacement. Geneva: Internal Displacement Monitoring Centre; 2021. Burns R. New frontiers of philanthro‐capitalism: digital technologies and humanitarianism. Antipode. 2019;51:1101-22. Aretz YL. Data philanthropy. SSRN Electron J. 2019. doi: 10.2139/ssrn.3320798. Taylor L. The ethics of big data as a public good: which public? Whose good? Philos Trans R Soc A Math Phys Eng Sci. 2016;374:20160126. Duffield M. The resilience of the ruins: towards a critique of digital humanitarianism. Resilience. 2016;4:147-65. McDonald S. Ebola: A Big Data Disaster [Internet]. Bengaluru and Delhi; 2016. Report No.: 2016.01. Available from: https://cis-india.org/papers/ebola-a-big-data-disaster. Blanchet K, Ramesh A, Frison S, Warren E, Hossain M, Smith J, et al. Evidence on public health interventions in humanitarian crises. Lancet. 2017;390:2287-96. Pham KT, Sattigeri P, Dhurandhar A, Jacob AC, Vukovic M, Chataigner P, et al. Real-time understanding of humanitarian crises via targeted information retrieval. IBM J Res Dev. 2017;61:7:1-7:12. Martin-Shields C. The technologists dilemma: ethical challenges of using crowdsourcing technology in conflict and disaster-affected regions. Georget J Int Aff. 2013;14:157-63. Phillips J. Risk in a digital age: understanding risk in virtual networks through digital response networks (DRNs). Int Dev Plan Rev. 2018;40:239-72. Meier P. Digital humanitarians : how big data is changing the face of humanitarian response. Boca Raton: CRC Press; 2014. OCHA. World humanitarian data and trends 2015. New York, NY: OCHA; 2015. Sapkota PP, Siddiqi K. Is ubiquitous technology for needs data management a game changer in humanitarian arena? Int J Inf Syst Crisis Response Manag. 2019;11:83-97. Madon S, Schoemaker E. Reimagining refugee identity systems: a sociological approach. In: Nielsen P, Kimaro HC, editors. Information and communication technologies for development Strengthening southern-driven cooperation as a catalyst for ICT4D. Cham: Springer International Publishing; 2019. p. 660-74. Eaton-Lee J, Shaughnessy E. Oxfam’s new policy on biometrics explores safe and responsible data practice. 2021. https://reliefweb.int/report/world/oxfam-s-new-policy-biometrics-explores-safe-and-responsible-data-practice. Accessed 30 Jun 2021. Additional Declarations No competing interests reported. Supplementary Files KreutzeretalScopingReviewEthicsHumanitarianDataAppendixA.docx Appendix A: Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) Checklist KreutzeretalScopingReviewEthicsHumanitarianDataAppendixB.docx Appendix B: Scoping Review Protocol Developed A Priori KreutzeretalScopingReviewEthicsHumanitarianDataAppendixC.docx Appendix C: Search Strategy and Keywords Used for Each Database KreutzeretalScopingReviewEthicsHumanitarianDataAppendixD.docx Appendix D: List of Studies Included in Scoping Review Results KreutzeretalScopingReviewEthicsHumanitarianDataAppendixE.docx Appendix E: Results Table for Studies Discussing Artificial Intelligence Cite Share Download PDF Status: Published Journal Publication published 15 Apr, 2025 Read the published version in BMC Medical Ethics → Version 1 posted Editorial decision: Revision requested 11 Jul, 2024 Reviews received at journal 10 Jul, 2024 Reviewers agreed at journal 02 Jul, 2024 Reviews received at journal 18 Jun, 2024 Reviewers agreed at journal 13 Jun, 2024 Reviewers invited by journal 13 Jun, 2024 Editor invited by journal 09 Apr, 2024 Submission checks completed at journal 09 Apr, 2024 Editor assigned by journal 09 Apr, 2024 First submitted to journal 05 Apr, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4224535","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":289497541,"identity":"03b1d338-596d-41af-82cf-12d46e5f651f","order_by":0,"name":"Tino Kreutzer","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7klEQVRIiWNgGAWjYDACZhBxAMarAIkwN+DVwQPXwgZinAFpYSSghQFZC2MbiEdAiz0778EHDGfuyRncb3/4uHJebTR/O1DLj4pteBzGl2zAcKPY2OAYQ7Lh2W3Hc2ccZmxg7DlzG48WHjMJhg8JiRuOMRyTbNx2LLcBqIWZsY2wlvoNxxjbfzbOOZY7nzgtNxISDI4xszE2NtTkbiCo5TCPsUHCmQTDmcfSmCUbjh3I3QjUchCfX9j7zxg++HAsQZ7v8PGHHxtq6nLnnT988MGPCtxawCABwTwMJg/gV48K6khRPApGwSgYBSMEAACeulg9Of7w0AAAAABJRU5ErkJggg==","orcid":"","institution":"Montreal Children's Hospital","correspondingAuthor":true,"prefix":"","firstName":"Tino","middleName":"","lastName":"Kreutzer","suffix":""},{"id":289497542,"identity":"2e58b9f3-810a-4fad-86db-2415f7515a61","order_by":1,"name":"James Orbinski","email":"","orcid":"","institution":"York University","correspondingAuthor":false,"prefix":"","firstName":"James","middleName":"","lastName":"Orbinski","suffix":""},{"id":289497543,"identity":"016eb416-3637-4352-abed-5fa79e7a08d7","order_by":2,"name":"Lora Appel","email":"","orcid":"","institution":"York University","correspondingAuthor":false,"prefix":"","firstName":"Lora","middleName":"","lastName":"Appel","suffix":""},{"id":289497544,"identity":"0bb5ba02-3f99-45d7-9d53-f7f7de1c9321","order_by":3,"name":"Aijun An","email":"","orcid":"","institution":"York University","correspondingAuthor":false,"prefix":"","firstName":"Aijun","middleName":"","lastName":"An","suffix":""},{"id":289497545,"identity":"5c345324-8c11-4162-a3b8-811acdc1b9af","order_by":4,"name":"Patrick Vinck","email":"","orcid":"","institution":"Harvard University","correspondingAuthor":false,"prefix":"","firstName":"Patrick","middleName":"","lastName":"Vinck","suffix":""}],"badges":[],"createdAt":"2024-04-05 18:59:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4224535/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4224535/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12910-025-01189-2","type":"published","date":"2025-04-15T15:57:01+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":54991715,"identity":"83ed4bd5-a03e-4676-98e5-dd680e07ab3c","added_by":"auto","created_at":"2024-04-19 17:22:56","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":160271,"visible":true,"origin":"","legend":"\u003cp\u003ePRISMA Diagram\u003c/p\u003e","description":"","filename":"Figure1PRISMAdiagram.png","url":"https://assets-eu.researchsquare.com/files/rs-4224535/v1/57d8c5e355110f2f002c46b5.png"},{"id":54991707,"identity":"da1bde02-25ac-47f4-a96a-d1c28ffc0812","added_by":"auto","created_at":"2024-04-19 17:22:52","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":513719,"visible":true,"origin":"","legend":"\u003cp\u003eMap Showing the Number of Authors per Country of Primary Affiliation\u003c/p\u003e","description":"","filename":"Figure2countryauthorlistmap.png","url":"https://assets-eu.researchsquare.com/files/rs-4224535/v1/882bc0a678141943a7027e68.png"},{"id":81050748,"identity":"f0fb9fa6-bf2f-4603-a9f4-38c78dc85c9c","added_by":"auto","created_at":"2025-04-21 16:03:21","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1807418,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4224535/v1/8a8f791b-024f-4f83-83c7-4d988bac62e8.pdf"},{"id":54991712,"identity":"16725ebe-d9b9-445a-958d-70362cecf6f0","added_by":"auto","created_at":"2024-04-19 17:22:55","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":55496,"visible":true,"origin":"","legend":"\u003cp\u003eAppendix A: Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) Checklist\u003c/p\u003e","description":"","filename":"KreutzeretalScopingReviewEthicsHumanitarianDataAppendixA.docx","url":"https://assets-eu.researchsquare.com/files/rs-4224535/v1/c69b3a3d25fdf626b43770de.docx"},{"id":54991714,"identity":"f48c1571-ab19-4334-8790-aca68c37709d","added_by":"auto","created_at":"2024-04-19 17:22:56","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":57280,"visible":true,"origin":"","legend":"\u003cp\u003eAppendix B: Scoping Review Protocol Developed A Priori\u003c/p\u003e","description":"","filename":"KreutzeretalScopingReviewEthicsHumanitarianDataAppendixB.docx","url":"https://assets-eu.researchsquare.com/files/rs-4224535/v1/9a0a6d101691cf48e2df6442.docx"},{"id":54991713,"identity":"f4e74d14-fccd-4213-aad5-1a2170cf5a7b","added_by":"auto","created_at":"2024-04-19 17:22:55","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":55292,"visible":true,"origin":"","legend":"\u003cp\u003eAppendix C: Search Strategy and Keywords Used for Each Database\u003c/p\u003e","description":"","filename":"KreutzeretalScopingReviewEthicsHumanitarianDataAppendixC.docx","url":"https://assets-eu.researchsquare.com/files/rs-4224535/v1/1a3847ab4e01b1cc571a4fe6.docx"},{"id":54991710,"identity":"447cfc46-7639-487f-9c84-506340ff8475","added_by":"auto","created_at":"2024-04-19 17:22:55","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":52842,"visible":true,"origin":"","legend":"\u003cp\u003eAppendix D: List of Studies Included in Scoping Review Results\u003c/p\u003e","description":"","filename":"KreutzeretalScopingReviewEthicsHumanitarianDataAppendixD.docx","url":"https://assets-eu.researchsquare.com/files/rs-4224535/v1/dacfaeae59fc878852a4e879.docx"},{"id":54991711,"identity":"d2599d80-ba87-4933-8cc0-c4aef525d4fe","added_by":"auto","created_at":"2024-04-19 17:22:55","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":50055,"visible":true,"origin":"","legend":"\u003cp\u003eAppendix E: Results Table for Studies Discussing Artificial Intelligence\u003c/p\u003e","description":"","filename":"KreutzeretalScopingReviewEthicsHumanitarianDataAppendixE.docx","url":"https://assets-eu.researchsquare.com/files/rs-4224535/v1/15cec9b160f832b436b397ef.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Ethical implications related to processing of personal data and artificial intelligence in humanitarian crises: A scoping review","fulltext":[{"header":"Background","content":"\u003cp\u003eOrganizations involved in providing humanitarian assistance work under strenuous circumstances and with limited funding to provide life-saving humanitarian assistance. However, resources for providing this assistance are far from sufficient. In 2022, donor governments provided US \u003cspan\u003e$\u003c/span\u003e 27\u0026nbsp;billion to help 216\u0026nbsp;million people in 69 countries\u0026mdash;a significantly smaller amount than the US \u003cspan\u003e$\u003c/span\u003e 51.7\u0026nbsp;billion required to assist all people in need of humanitarian assistance for that year [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. This considerable shortfall highlights the urgent need to better assess humanitarian needs and to do so at minimal cost.\u003c/p\u003e \u003cp\u003eThe aim of this review is to map the range of ethical issues that have been raised in the academic literature regarding data processing of people affected by humanitarian crises. Humanitarian organizations rely on processing increasingly large amounts of data to inform their operations, much of which is collected directly from affected populations (e.g., through registrations, household surveys, or cash disbursements). At the same time, the people working for these organizations have themselves often become targets of kidnappings and killings, which has led organizations to increasingly resort to remote methods of managing operations and collecting data from affected people [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. The COVID-19 pandemic has accelerated this trend of the increased use of remote methods. This combination of factors has led to an exponential increase in the amount of personal data that is being distributed, stored, and analyzed in various locations around the world. At the same time, humanitarian organizations are continuously seeking innovations involving information and communication technologies (ICT) in the pursuit of operational gains in effectiveness and efficiency. This practice is expected to accelerate further with the growing availability and sophistication of artificial intelligence (AI) technologies in the health and humanitarian sectors.\u003c/p\u003e \u003cdiv id=\"Sec2\" class=\"Section2\"\u003e \u003ch2\u003eDefinitions\u003c/h2\u003e \u003cp\u003eA review by Schofield et al. [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e] found that the vast majority of included studies discussing \u0026ldquo;ethical challenges\u0026rdquo; in healthcare had failed to include an explicit definition of how that term was understood by the respective authors, leading to potential misunderstandings and ambiguity. This section, therefore, will first provide working definitions for the key terms and concepts discussed in this study. Humanitarian assistance is understood here to refer to coordinated actions that save lives and alleviate suffering of crisis-affected populations [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. It also includes \u0026ldquo;protection\u0026rdquo;, which \u0026ldquo;encompasses all activities aimed at obtaining full respect for the rights of the individual in accordance with the letter and the spirit of the relevant bodies of law\u0026rdquo; [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Humanitarian crises are defined here as a \u0026ldquo;series of events representing a critical threat to the health, safety, security or wellbeing of a community, usually over a wide area\u0026rdquo; [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. For the purposes of this study, data processing is understood as: \u0026ldquo;Any operation or set of operations which is performed on data or on sets of data, whether or not by automated means, such as collecting, registering, storing, adapting or altering, cleaning, filing, retrieving, using, disseminating, transferring and retaining or destroying\u0026rdquo; [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Ethical issues are defined in this study as actions that may not conform to moral standards, particularly those set out by various humanitarian principles [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e] because of the risks they present.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eContext\u003c/h3\u003e\n\u003cp\u003eOrganizations turning to new or existing digital tools to collect, store, or analyze data more efficiently may knowingly or inadvertently introduce new ethical issues affecting people who are already vulnerable [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Weighing the responsible use of new technologies in humanitarian crises is fraught with a number of ethical issues [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] that are increasingly being highlighted in specific circumstances such as refugee registrations [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e], health emergency response [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], or the use of drones in humanitarian assistance [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. In practice, ethical decisions are made\u0026mdash;knowingly or unknowingly\u0026mdash;on a daily basis about what data to collect, which tools to use, or how and with whom to share this information to avoid adverse consequences [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. In light of such challenges, organizations rarely choose to forego new tools altogether, such as Oxfam\u0026rsquo;s decision in 2015 to halt the use of biometrics in its programs in order to assess the potential risks [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Rather, some organizations are more likely to invest in new innovations without considering, weighing, or fully grasping the long-term ethical issues [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eHowever, because of these challenges, more guidelines are now being produced for the ethical processing of data for humanitarian assistance purposes, with the goal of minimizing or eliminating risks to vulnerable people. Notable examples include \u003cem\u003eData Responsibility in Humanitarian Action\u003c/em\u003e by the Inter-Agency Standing Committee [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e], the \u003cem\u003eHandbook on Data Protection in Humanitarian Action\u003c/em\u003e by the International Committee of the Red Cross [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e], and the \u003cem\u003eData Responsibility Guidelines\u003c/em\u003e by the United Nations Office for the Coordination of Humanitarian Affairs [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Similarly, regulatory environments are changing in many countries (such as the European Union\u0026rsquo;s General Data Protection Regulation, GDPR), which have moved many humanitarian organizations to change their approaches to data processing in order to improve data privacy. Focusing on the issue of ethical design of new tools, Krishnaraj et al. [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e] have created practical guidelines that aim to mitigate risks as early as possible [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. But the speed of technological innovation means that such guidance can quickly become out of date as new data technology tools appear and organizations respond to new circumstances (such as insecure environments or lack of access to populations during the COVID-19 pandemic). Artificial intelligence (AI) systems that use machine learning and other methods for automating data processing may usher in a completely new set of ethical issues that humanitarian organizations will have to confront [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Innovations using AI in the medical and health sectors has been growing significantly for years and is showing important premises, such as in the discovery of new classes of antibiotics [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. At the same time, large language models such as ChatGPT [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e] that excel at generating and summarizing human language are generating yet another set of novel ethical issues [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e], including in the health and medical sectors [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Although a considerable number of studies discuss the ethics of using various technologies in humanitarian assistance, to date, there is little evidence that there has been a comprehensive review of relevant ethical issues in the published literature.\u003c/p\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eHumanitarian and Technical Nomenclatures\u003c/h2\u003e \u003cp\u003eConducting this type of review is challenging due to the wide-ranging nature of humanitarian assistance, lack of well-defined nomenclature for data processing technologies and activities, and that relevant research may be published in the intersecting fields of ethics research, design, engineering, health, medicine, geography, development, social science, and technology research, among others. Previous scoping reviews focusing on humanitarian assistance only addressed more limited contexts, such as natural disasters [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e], and displaced populations [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e], or did not include terms to capture more novel humanitarian activities such as responding to large-scale migration or public health emergencies [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Other relevant studies in the past have included a literature review focused on social media and privacy issues (based on literature published between 2013 and 2014 [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e], a scoping review on the types of digital tools used during the 2014\u0026ndash;2016 Ebola outbreak in West Africa [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e], a scoping review on the impact of health-related tools in humanitarian crises [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e], and a scoping review on ethical considerations related to the use of drones in humanitarian assistance [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. However, we have not found a sufficiently comprehensive sets of keywords that could be used to search databases for any of this study\u0026rsquo;s three inclusion criteria (people affected by humanitarian crises, processing data for humanitarian assistance, meaningful discussion of ethical issues). Even though the data collection for this study was conducted in 2020, it remains the most recent scoping review on this topic.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eEthical Frameworks\u003c/h2\u003e \u003cp\u003eAnother distinct challenge is the lack of established ethical categories or theories used by studies discussing ethics in the humanitarian sector [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. First introduced by the International Federation of Red Cross and Red Crescent Societies [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e], the four humanitarian principles (humanity, impartiality, independence, and neutrality) are now widely used among many humanitarian organizations [see, for example, 38], in international law [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e], as well as in ethical codes attempting to guide the actions of the humanitarian sector as a whole [see, e.g., 40, 41]. However, previous studies have shown the difficulty of applying these humanitarian principles in everyday practice [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e], in guiding the use of information technology [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e], or in mapping humanitarian organizations\u0026rsquo; ethical obligations [\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. In particular, the broad \u003cem\u003ehumanity\u003c/em\u003e \u0026ldquo;principle\u0026rdquo; has been argued as being better understood as an absolute moral value rather than an ethical principle [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eDeveloped in parallel, the ethical value categories of \u003cem\u003eautonomy\u003c/em\u003e, \u003cem\u003ebeneficence\u003c/em\u003e, \u003cem\u003enon-maleficence\u003c/em\u003e, and \u003cem\u003ejustice\u003c/em\u003e, are widely used in the fields of bioethics and research ethics, and have been defined and discussed in their application in detail by Beauchamp and Childress [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. A definition of each ethical value category is provided in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. The four categories reflect work largely in the decades following World War II that have aimed to better protect research participants, including the Nuremberg Code [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e], the 1964 Declaration of Helsinki [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e] and the Belmont Report [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e].\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDefinitions of Each Ethical Value Category, Based on Beauchamp and Childress [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePrinciple\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDefinition\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRespect for autonomy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRespecting the decision-making capacities of autonomous persons\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBeneficence\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eProviding benefits and balancing benefits against risks and costs\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNon-maleficence\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAvoiding the causation of harm\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJustice\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDistributing benefits, risks, and costs fairly\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eAs a result of the challenges in applying humanitarian ethical principles, a growing number of studies use the four ethical value categories as a better operational ethical terms to reference ethical issues inherent to humanitarian practice [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]. We also chose to employ these four ethical value categories to group the ethical issues identified in the literature and to better link the nascent field of humanitarian ethics to the larger theoretical and practical advances in the fields of bioethics and research ethics.\u003c/p\u003e \u003cp\u003eThe aim of this review is to map the range of ethical issues that have been raised in the academic literature regarding processing relevant to people affected by humanitarian crises. This study contributes to the existing academic discussion in three important ways. First, this study presents the first comprehensive review of the ethical considerations in processing data from individuals affected by humanitarian crises, addressing a significant gap in the literature. Second, it addresses the challenges of fragmented terminology by establishing an evidence-based search strategy to cover topics in the intersection of humanitarian assistance, data processing, and ethical implications. Third, the study introduces a clear, transparent framework for defining what constitutes a \u0026ldquo;humanitarian crisis,\u0026rdquo; providing a consistent basis for the inclusion or exclusion of different studies, which may help avoid subjective biases in research selection for future studies.\u003c/p\u003e \u003c/div\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec7\"\u003e\n \u003ch2\u003eStudy Protocol\u003c/h2\u003e\n \u003cp\u003eWe chose to conduct a scoping review as this method is best suited for generating a broad overview of relevant evidence, to examining emerging areas of research, to clarifying key concepts, and to identifying gaps in the literature [\u003cspan\u003e51\u003c/span\u003e]. A study protocol was developed prior to data collection and screening using the scoping review method established by Arksey and O\u0026rsquo;Malley [\u003cspan\u003e52\u003c/span\u003e] as further refined by Levac et al. [\u003cspan\u003e53\u003c/span\u003e] and follows the framework maintained by the Joanna Briggs Institute [\u003cspan\u003e54\u003c/span\u003e]. The protocol was revised based on feedback received from the research team and incorporated the results from a pilot conducted for this study November-December, 2019. It follows the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) reporting guidelines [\u003cspan\u003e55\u003c/span\u003e]. The final version of the PRISMA-ScR checklist and the study protocol are available in Appendix A and Appendix B, respectively.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec8\"\u003e\n \u003ch2\u003eIdentifying the Research Question\u003c/h2\u003e\n \u003cp\u003eThe specific research questions of this scoping review were:\u003c/p\u003e\n \u003cp\u003e1. Which ethical issues have been raised in the literature related to processing data from people affected by humanitarian crises in order to inform humanitarian assistance?\u003c/p\u003e\n \u003cp\u003e2. To what extent do real-world examples of ethical issues reflect the concerns presented in the literature?\u003c/p\u003e\n \u003cp\u003e3. Which technologies were the focus of concern over these ethical issues?\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec9\"\u003e\n \u003ch2\u003eEligibility Criteria\u003c/h2\u003e\n \u003cp\u003eThe following eligibility criteria for the selection of relevant studies were established \u003cem\u003ea priori\u003c/em\u003e as per the categories and requirements for scoping review protocols [\u003cspan\u003e55\u003c/span\u003e].\u003c/p\u003e\n \u003cul\u003e\n \u003cli\u003e\n \u003cp\u003e\u003cem\u003eCondition/Domain\u003c/em\u003e: Ethical issues stemming from the processing of data relating to people affected by a humanitarian crisis with the explicit goal or potential of informing humanitarian assistance.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003e\u003cem\u003ePopulation\u003c/em\u003e: People affected by a humanitarian crisis, including armed conflicts, natural disasters, and large public health emergencies, as well as refugees and transborder migrants fleeing from such a crisis\u0026mdash;regardless of their current location. We also included studies that concern humanitarian assistance (including related fields such as disaster response or emergency management) that are global in scope. Studies about natural disasters were only included if the study focused on events in low or lower middle-income countries (defined as countries that ranked low income or low middle income at least once by the World Bank between 2011\u0026ndash;2020) [\u003cspan\u003e56\u003c/span\u003e]. The Ebola outbreak in West Africa (2014\u0026ndash;2016) was included as it was widely considered to be a humanitarian crisis in scope [\u003cspan\u003e57\u003c/span\u003e]. We used the Financial Tracking Service by UN OCHA [\u003cspan\u003e58\u003c/span\u003e] to judge if an event should be considered a humanitarian crisis (defined as whether a given country was a recipient of humanitarian aid in the same year).\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003e\u003cem\u003eInterventions\u003c/em\u003e: Data processing relating to people affected by a humanitarian crisis with the explicit goal or potential of informing humanitarian assistance. Excluded were studies that focus on technologies that do not process data on affected people, such as robotics for clearing debris or land mines, algorithmic models for predicting the occurrence or impacts of natural hazards, or tools used for planning humanitarian logistics (e.g., relief/distribution networks, supply chain management, and resource scheduling).\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003e\u003cem\u003eOutcomes\u003c/em\u003e: Studies that investigate ethical issues stemming from the processing of data (as defined above) were included only if they contained a significant discussion about this subject. During the screening stage, studies were eligible for inclusion if the abstract referenced or mentioned potential ethical issues. During the full text review this was assessed qualitatively by two reviewers.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003e\u003cem\u003eStudy Designs\u003c/em\u003e: All study designs were eligible for inclusion, including empirical studies, commentaries, and theoretical papers. Excluded were non-peer reviewed studies as well as book reviews. In order to establish a robust foundation for formulating evidence-based recommendations and for feasibility reasons, this research was limited peer-reviewed literature.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003e\u003cem\u003eContext\u003c/em\u003e: For feasibility reasons, we restricted the review to studies published after 1 January 2010.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003e\u003cem\u003eSetting\u003c/em\u003e: Studies in all countries or territories affected by a humanitarian crisis (or relevant host countries for refugee or cross-border migrant or displaced populations) were included, as defined above.\u003c/p\u003e\n \u003c/li\u003e\n \u003c/ul\u003e\n \u003cdiv id=\"Sec10\"\u003e\n \u003ch2\u003eSearch Strategy and Information Sources\u003c/h2\u003e\n \u003cp\u003eComprehensive literature searches of electronic databases were conducted on 31 March 2020 for studies published between 2010 and 2019, using Ovid, Ebsco, Web of Science, and Proquest to search 20 databases for relevant studies. Only studies published in English, French, or Spanish were included.\u003c/p\u003e\n \u003cp\u003eAs recommended by the scoping review guidelines described above, keywords were selected and piloted in multiple iterations to identify all relevant articles. We had previously identified 34 studies, and these were used as a minimum search target. After an initial search showed that only 13 were included, we repeated the database search over several iterations with additional terms until all 34 studies were reflected in the results. This yielded additional keywords such as \u0026ldquo;risks\u0026rdquo; and \u0026ldquo;challenges\u0026rdquo; to represent ethical challenges, as well as \u0026ldquo;innovation\u0026rdquo; and \u0026ldquo;experimentation\u0026rdquo; which are sometimes used to refer to data processing activities. Further, careful searching for terms such as \u0026ldquo;acute malnutrition\u0026rdquo; or \u0026ldquo;forcibly displaced population\u0026rdquo; were also found to describe specific phenomena in a humanitarian crisis without using terms such as \u0026ldquo;refugees\u0026rdquo; or \u0026ldquo;humanitarian\u0026rdquo; in the study\u0026rsquo;s metadata. Likewise, to find all studies that discuss processing data of affected people, we iteratively expanded our search terms to include specific technologies (e.g., biometrics, remote sensing), emerging practices (e.g., remote management, crowdsourcing), or shorthand keywords introduced by researchers (e.g., experimentation, crisis informatics, innovation). A sample of the search strategy for the Ovid databases is displayed in Table\u0026nbsp;\u003cspan\u003e2\u003c/span\u003e. The complete search syntax for each database can be found in Appendix C.\u003c/p\u003e\n \u003cdiv\u003e\n \u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv\u003eTable 2\u003c/div\u003e\n \u003cdiv\u003e\n \u003cp\u003eSearch Strategy for Ovid Databases\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eConcept\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eKeyword and syntax\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"16\"\u003e\n \u003cp\u003eHumanitarian assistance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ehumanitarian*.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003erelief work.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eaid work.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e(disaster? Adj (relief or response? Or assistance)).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eemergency relief.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e((conflict? Or war?) adj10 (human rights or public health)).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e(ebola adj6 (west africa or sierra leone or liberia or guinea or 2014 or 2013)).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eacute malnutrition.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e(refugee* adj2 (camp* or assistance or population?)).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e(displace* adj2 (forced or forcibly or population? Or human? Or internal*)).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e(((population? Or person* or communit*) adj3 affected) adj1 (conflict? Or violence)).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eor/ 1\u0026ndash;11\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e(cris?s or emergenc* or disaster?).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ehumanitarian*.af.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13 and 14\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12 or 15\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"24\"\u003e\n \u003cp\u003eICT for data collection\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eict.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003etechnolog*.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e((data or information) adj2 (system* or manage* or collection or analys?s or process*)).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e(blockchain or distributed ledger).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e(ai or artificial intelligence or machine learning or algorithm*).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ebiometric*.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003esmartphone app*.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eremote sensing.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eanalytics.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003edigital*.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eexperimentation.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eautomat*.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003einnovation?.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eremote management.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ecyber.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e32\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ebig data.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e(sms or text messag* or interactive voice recognition or online survey*).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e(kobotoolbox or kobo or odk or open data kit).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ecrowdsource*.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003esocial media.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ecrisis adj (informatics or data or map*).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e38\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003edigiti?ation.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003edatafication.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eor/ 17\u0026ndash;39\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"18\"\u003e\n \u003cp\u003eEthical concerns\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003econcern?.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003erisk?.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003echallenge?.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eharm?.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eprivacy.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eprotection?.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ehumanitarian adj (principle? Or standard? Or guideline?).tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eproblem?.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e49\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ebias?.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eethic*.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003econsequence?.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e52\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ecritique?.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e53\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003einsecurity.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e54\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eimplications.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e55\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eperil?.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eimpact?.tw.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e57\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eor/ 41\u0026ndash;56\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e16 and 40 and 57\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec11\"\u003e\n \u003ch2\u003eStudy Selection\u003c/h2\u003e\n \u003cp\u003eStudy selection and coding were done using the DistillerSR systematic review software [\u003cspan\u003e59\u003c/span\u003e]. Using the \u003cem\u003ea priori\u003c/em\u003e eligibility criteria, we developed questionnaires for selecting citations during discrete title, abstract, and full text review stages. Two reviewers independently selected studies during each screening stage.\u003c/p\u003e\n \u003cp\u003eRegular meetings to discuss rating discrepancies and to compare working definitions were held during the review of the first 1,000 references in the title screening stage and for the first 100 references during the abstract screening stage. Any conflicts during the title and abstract screening stages were included in the full text review. In the full text screening stage, daily meetings were held during the review of the first 20 references to discuss rating discrepancies and to improve working definitions of terms. Rating discrepancies were resolved by discussion, and in five cases, by using a third adjudicator.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\"\u003e\n \u003ch2\u003eData Collection Process\u003c/h2\u003e\n \u003cp\u003eFor included studies, we extracted details on study characteristics (year of publication, countries of all authors, author organization types), population characteristics (type of humanitarian crisis), intervention characteristics (purpose of data processing, technologies described), and outcomes (specific ethical issues identified, whether studies used real-world examples to identify issues). Author organization types were coded for all listed affiliations, while author country was extracted only from the first-listed affiliation. For each country, we additionally tabulated the geographic region and income level, using the 2020 World Bank classification scheme [\u003cspan\u003e56\u003c/span\u003e].\u003c/p\u003e\n \u003cp\u003eThe data extraction form was created in the DistillerSR software. It was then piloted based on a random sample of 10 included studies and modified based on discussions and feedback from the two reviewers. As per the study protocol, since the number of included citations was greater than 30, data extraction was done by one reviewer and verified by another. The data extraction form included several pre-coded ethical issues, but additional emergent issues could be entered qualitatively in text format.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\"\u003e\n \u003ch2\u003eSynthesis\u003c/h2\u003e\n \u003cp\u003eWe summarized results quantitatively (using frequencies) and qualitatively (using descriptive analytics). We analyzed and coded the ethical issues related to data processing that were entered in text form using SPSS 25. Specific issues described by authors could be assigned to one or more categories of ethical issues. Issue codes were updated iteratively and recursively by creating new codes based on new observations and through constant retrospective reviews of previously collected data. In some cases, rarely-mentioned codes were also merged retrospectively to limit the size of the final list of issues. The ethical issues mentioned in each study were then grouped into the ethical value categories of autonomy, beneficence, non-maleficence, and justice, based on which category was deemed to be the affected most.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eLiterature Search\u003c/h2\u003e \u003cp\u003eThe database literature search returned 8,387 citations (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). After removing duplicates, 5,999 were included for screening. 3,951 were excluded during the title screening stage and 1,752 during abstract screening. After reviewing full texts of 296 potentially relevant studies, 198 were excluded. As a result, 98 were included in this scoping review (full list of citations listed in Appendix D).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eStudy Characteristics\u003c/h2\u003e \u003cp\u003eThe included 98 studies were published between 2010 and 2019, as shown in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. The majority (n\u0026thinsp;=\u0026thinsp;72) were published after 2015, and the most common publication year was 2019 (n\u0026thinsp;=\u0026thinsp;28). Most were written by authors based in Europe and Central Asia (n\u0026thinsp;=\u0026thinsp;55) and North America (n\u0026thinsp;=\u0026thinsp;37), while only a small number of studies included authors from East Asia \u0026amp; Pacific (n\u0026thinsp;=\u0026thinsp;8), South Asia (n\u0026thinsp;=\u0026thinsp;4), Sub-Saharan Africa (n\u0026thinsp;=\u0026thinsp;3), Middle East and North Africa (n\u0026thinsp;=\u0026thinsp;3), and Latin America and the Caribbean (n\u0026thinsp;=\u0026thinsp;2), as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. 31 studies included an author from the United States while about one quarter (n\u0026thinsp;=\u0026thinsp;27) included an author from the United Kingdom. Overall, 92 studies included at least one author from a high-income country while a smaller number included at least one author from an upper middle-income country (n\u0026thinsp;=\u0026thinsp;7) or lower middle-income country (n\u0026thinsp;=\u0026thinsp;7). No study included an author from a low-income country. Similarly, no study included an author from China. The vast majority (n\u0026thinsp;=\u0026thinsp;90) of studies included at least one author from an academic institution, while only 7 studies included at least one author affiliated with a humanitarian organization.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eStudy Characteristics (n\u0026thinsp;=\u0026thinsp;98)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharacteristic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCount (%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eYear of publication\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2010\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2 (2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2012\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1 (1%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2013\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5 (5%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2014\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9 (9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2015\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9 (9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2016\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20 (20%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2017\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11 (11%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2018\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e13 (13%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2019\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e28 (29%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eRegion represented by authors\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEurope \u0026amp; Central Asia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e55 (56%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNorth America\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e37 (38%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEast Asia \u0026amp; Pacific\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8 (8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSouth Asia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4 (4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSub-Saharan Africa\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 (3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMiddle East \u0026amp; North Africa\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 (3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLatin America \u0026amp; Caribbean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2 (2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eCountry income level based on author location\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHigh income\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e92 (94%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUpper middle-income\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7 (7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLower middle-income\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7 (7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLow-income\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eParent organization type based on author affiliation\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAcademic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e90 (92%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFor-profit\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12 (12%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNon-profit\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7 (7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHumanitarian\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7 (7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eThink tank\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2 (2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGovernment\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1 (1%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eType of Humanitarian Crisis\u003c/h2\u003e \u003cp\u003eSimilar numbers of studies focused on or included examples of natural disasters and armed conflict (n\u0026thinsp;=\u0026thinsp;37 and n\u0026thinsp;=\u0026thinsp;35, respectively), as shown in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e. Of the 98 studies selected, 31 discussed people displaced by a humanitarian crisis, whereas 19 focused on large public health emergencies. Twenty studies were general in nature and only discussed the fields of humanitarian assistance, emergency management, or disaster response without providing specific examples.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eTypes of Humanitarian Crises Discussed (n\u0026thinsp;=\u0026thinsp;98)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eType of humanitarian crisis\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCount (%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNatural disaster\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e37 (38%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eArmed conflict\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e35 (36%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePeople displaced by a humanitarian crisis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31 (32%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLarge public health emergency\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18 (18%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot specified\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e19 (19%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003ePurpose of Data Processing\u003c/h2\u003e \u003cp\u003eWhile most studies reported more than one purpose, the most common data processing purpose was conducting assessments (n\u0026thinsp;=\u0026thinsp;35), such as needs assessments or damage surveys (see Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Twenty-four studies examined different forms of case management (e.g., refugee registrations), while 18 discussed handling of medical or public health data. Twenty-three did not specify any reasons for data processing but instead discussed in theoretical terms the use of information and communication technologies or data processing in humanitarian assistance.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eData Processing Purposes and Technologies Described by Studies (n\u0026thinsp;=\u0026thinsp;98)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePurposes of data processing\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCount (%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAssessment (of needs, damage, etc.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e35 (36%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRegistration / case management\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e24 (24%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eForecasting / modeling / early warning\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18 (18%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMedical care or public health\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18 (18%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDelivery of assistance\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12 (12%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAccountability (complaints, feedback collection, etc.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9 (9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCash transfer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8 (8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLogistics\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8 (8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSearch and rescue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8 (8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHuman rights violations\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7 (7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOther\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e17 (17%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot specified\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e23 (23%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eTechnologies Described\u003c/h2\u003e \u003cp\u003eThe most commonly described technologies used for data processing were social media (discussed by 47 studies), crowdsourcing (n\u0026thinsp;=\u0026thinsp;44), various forms of mapping and other forms of geographic information systems (GIS; n\u0026thinsp;=\u0026thinsp;42), whereas nearly one in three studies focused on big data (n\u0026thinsp;=\u0026thinsp;29) or private messaging services (n\u0026thinsp;=\u0026thinsp;29).\u003c/p\u003e \u003cp\u003eOne quarter (n\u0026thinsp;=\u0026thinsp;25) discussed various forms of tools and technologies related to artificial intelligence (AI), including the use of algorithms and machine learning. The capture of satellite images for humanitarian assistance and the collection of biometrics (typically, fingerprint or iris scans) were discussed by 22 and 21 studies, respectively. Other technologies cited are shown in Table\u0026nbsp;\u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003e6\u003c/span\u003e. Five studies did not discuss any specific technologies used for data processing.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab6\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 6\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eTechnologies Described by Studies (n\u0026thinsp;=\u0026thinsp;98)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSpecific technologies described\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCount (%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSocial media\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e47 (48%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCrowdsourcing\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e44 (45%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMapping / GIS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e42 (43%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBig data\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e29 (30%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSMS or private messaging software\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e29 (30%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAI / algorithms / machine learning\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e25 (26%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSatellite imagery\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e22 (22%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBiometrics\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21 (21%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInformation systems\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21 (21%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDrones / UAV\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18 (18%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCash distribution\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11 (11%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMedical data\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10 (10%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCall data records\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9 (9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eData storage\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5 (5%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBlockchain / distributed ledger technology\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4 (4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eComputer-assisted personal interviewing\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 (3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot specified\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5 (5%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eEthical Issues Identified\u003c/h2\u003e \u003cp\u003eAs shown in Table\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e7\u003c/span\u003e, we identified 22 ethical issues in the studies under investigation, which were grouped according to the four previously identified bioethical values categories. Eight issues were attributed to the ethical value category of autonomy, six to beneficence, seven to non-maleficence, and five to justice. On average, studies cited seven different ethical issues each (\u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;6.86, \u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;3.32), ranging from more than two for non-maleficence issues (\u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.43, \u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.3) to less than one for justice issues (\u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.96, \u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.76; see Table\u0026nbsp;\u003cspan refid=\"Tab8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). The vast majority of studies mentioned issues related to non-maleficence (n\u0026thinsp;=\u0026thinsp;91) and beneficence (n\u0026thinsp;=\u0026thinsp;87). Slightly fewer studies discussed issues concerning justice (n\u0026thinsp;=\u0026thinsp;71) and autonomy (n\u0026thinsp;=\u0026thinsp;69).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab7\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 7\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eEthical Issues Identified (n\u0026thinsp;=\u0026thinsp;98)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eEthical issues identified\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCount (%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"7\" rowspan=\"8\"\u003e \u003cp\u003eAutonomy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLack of consent: Data is collected without informed consent\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e50 (51%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eData agency: People do not have the right to control, access, or delete their data\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e30 (31%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLack of respect: People/communities are not treated with respect\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e29 (30%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAutonomy: Unwillingness to share data does not lead to disadvantages (e.g., exclusion from assistance or protection\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e24 (24%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eParticipation: People/communities are not involved in decisions to use of new/experimental technologies for collecting data\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e16 (16%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUndisclosed use: Data may be used beyond purposes for which they were collected\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8 (8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLack of group agency: Processed information is not available to affected communities\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6 (6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003eAny implication related to Autonomy\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e69 (70%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e \u003cp\u003eBeneficence\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUnreliability: Processed data is inaccurate and does not sufficiently reflect reality to inform assistance\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e60 (61%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDependence: Data is processed with the assistance of a political, economic, or military entity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e47 (48%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLack of action: Processed data is not utilized to inform assistance to the affected person/community\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e38 (39%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNon-neutrality: Data is processed in a way that benefits or appears to benefit one side of the conflict over the other\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e28 (29%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIneffective or inefficient: Not producing expected result, unmet expectations\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4 (4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003eAny implication related to Beneficence\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e87 (89%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"6\" rowspan=\"7\"\u003e \u003cp\u003eNon-maleficence\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePrivacy: Personal/sensitive data is shared with third parties\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e72 (73%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHarm: People suffer physical or psychological harm as a result of data processing\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e66 (67%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eData security: Personal/sensitive data is not protected against malicious actors\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e41 (42%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePower imbalance: Data processing reinforces or worsens a lack of power of affected people\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e33 (34%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eExcess: More data was collected than necessary\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e17 (17%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRedress/rectification: People do not have the ability to correct wrong information about them or receive compensation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9 (9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003eAny implication related to Non-maleficence\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e91 (93%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"4\" rowspan=\"5\"\u003e \u003cp\u003eJustice\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBias: Data is processed in a way that may (dis)advantage some people disproportionate to their humanitarian needs\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e61 (62%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLack of accountability: Endangering (or not protecting) rights; absolving responsibility\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e16 (16%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUnequal access to technology / exclusion from data collection\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e14 (14%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUnfair distribution of risks and benefits\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3 (3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003eAny implication related to Justice\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e71 (72%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e[insert Table\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e7\u003c/span\u003e here]\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab8\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 8\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eNumber of Ethical Issues Cited by Ethical Value Category (n\u0026thinsp;=\u0026thinsp;98)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBioethical value category\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003emean, SD (min to max)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eAll ethical value categories\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003e6.97, 3.4 (1 to 15)\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAutonomy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.73, 1.59 (0 to 7)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBeneficence\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.82, 1.16 (0 to 5)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNon-maleficence\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.44, 1.3 (0 to 6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJustice\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.98, 0.77 (0 to 3)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe most frequently cited ethical issue categorized under the value category of autonomy was data being collected without sufficient informed consent (n\u0026thinsp;=\u0026thinsp;52). For example, Shoemaker et. al. [\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e] found through interviews with refugees are frequently being asked by humanitarian organizations to provide personal information that the respondents considered intrusive, without being offered a justification on why this was relevant.\u003c/p\u003e \u003cp\u003eWithin the value category of beneficence, the ethical issue most frequently mentioned by studies was processed data being inaccurate and not sufficiently reflecting reality to inform assistance (n\u0026thinsp;=\u0026thinsp;61). This is illustrated by Paul and Sosale [\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e], who cite the challenges of using social media as a basis to inform humanitarian assistance. In an example the authors cite, the same information was re-posted multiple times by well-meaning users, making it difficult for emergency responders after a severe flooding event to identify new information that might require a team to be dispatched.\u003c/p\u003e \u003cp\u003eFalling under the value category of non-maleficence, the most-cited ethical issue (n\u0026thinsp;=\u0026thinsp;74) were privacy concerns in cases where personal or sensitive data may be shared with third parties. For example, Hayes and Kelly [\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e] discuss how personal requests for help that are aggregated by a crowdsourcing platform such as Ushahidi can make personal information publicly available, including to bad actors trying to exploit vulnerable people.\u003c/p\u003e \u003cp\u003eThe most frequently mentioned ethical issue categorized under the justice value category was biased data processing leading to (dis)advantaging people disproportionate to their humanitarian needs (n\u0026thinsp;=\u0026thinsp;63). This issue, which often relates to different forms of sampling problems that could endanger the impartial distribution of aid, has become more pressing as more organizations turn to \u0026ldquo;big data\u0026rdquo; solutions for informing humanitarian assistance without properly understanding their limitations [\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eInformation Sources for Ethical Issues\u003c/h2\u003e \u003cp\u003eAs shown in Table\u0026nbsp;\u003cspan refid=\"Tab9\" class=\"InternalRef\"\u003e9\u003c/span\u003e, slightly over half of studies (n\u0026thinsp;=\u0026thinsp;52) cited at least one real-world example of an ethical issue, usually based on anecdotal information found in news reports or other published literature [see, e.g., 64, 65]. Fully 29 studies included ethical issues that were raised by interviews or other kinds of consultations with experts. Examples here include Shoemaker et. al. [\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e], who conducted qualitative interviews with 198 refugees in Lebanon, Jordan, and Uganda, as well as Vannini et al. [\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e], who interviewed nine representatives from organizations assisting transborder migrants in the United States. Four studies included a systematic review of the literature [\u003cspan additionalcitationids=\"CR34\" citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e].\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab9\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 9\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eInformation Sources of Ethical Issues (n\u0026thinsp;=\u0026thinsp;98)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSources of ethical issues\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCount (%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSpecific instances of ethical issue (rooted in real-life experience)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e52 (53%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEthical concerns raised in interviews/expert consultations\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e29 (30%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSystematic review of the literature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4 (4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eKey Results for Studies Discussing AI\u003c/h2\u003e \u003cp\u003eOf the 25 studies that discuss the use of AI, all were published since 2014, with about half (n\u0026thinsp;=\u0026thinsp;13) published after 2017 (see Appendix E for all figures related to the AI-related studies). The most common type of humanitarian crisis discussed in the 25 studies was natural disaster (n\u0026thinsp;=\u0026thinsp;9), followed by armed conflict and large public health emergencies (n\u0026thinsp;=\u0026thinsp;7, respectively). The most common purposes for data processing were related to assessments (n\u0026thinsp;=\u0026thinsp;9) as well as handling medical or public health data (n\u0026thinsp;=\u0026thinsp;7). The majority of the 25 studies related to AI also discussed big data (n\u0026thinsp;=\u0026thinsp;15), social media (n\u0026thinsp;=\u0026thinsp;15), and GIS (n\u0026thinsp;=\u0026thinsp;13). The vast majority mentioned ethical implications related to privacy (n\u0026thinsp;=\u0026thinsp;23) and the risk of physical or psychological harm (n\u0026thinsp;=\u0026thinsp;22). Eighteen studies related to data being processed in a way that may result in aid being distributed disproportionately with regard to people\u0026rsquo;s actual needs.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe aim of this review was to map the range of ethical issues that have been raised in the academic literature regarding data processing of people affected by humanitarian crises. This review identified 22 such ethical issues. Issues related to the value category of non-maleficence were brought up by the vast majority of studies (n\u0026thinsp;=\u0026thinsp;93), which dovetails with a strong trend in the recent literature focusing on the imperative of \u0026ldquo;do no harm\u0026rdquo; in humanitarian assistance [\u003cspan additionalcitationids=\"CR69\" citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e]. The risk of increasing harm (whether physical or psychological) as a result of data processing was mentioned by a high number of studies (n\u0026thinsp;=\u0026thinsp;68).\u003c/p\u003e \u003cp\u003ePrivacy concerns were cited by 74 studies\u0026mdash;far more commonly than all other issues\u0026mdash;reflecting an increased awareness of this issue over the last years across organizations and the media. Among studies discussing AI as a data processing technology, privacy concerns were even more prevalent, with 23 out of 25 (92%) mentioning this issue. This points to a significant worry across the humanitarian sector about the many ways in which personal data from affected people is being processed in a manner that may endanger their right to privacy, which is enshrined in the 1948 Universal Declaration of Human Rights [\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e]. Studies discuss a wide gamut of how personal privacy can be violated, including accidental or intentional sharing with third parties beyond what the affected person had agreed to during personal interviews, or if at all. Even in cases where informed consent was given, interviewees in vulnerable situations\u0026mdash;or who lack understanding of sophisticated data management, access and processing\u0026mdash;may not understand all the potential ways in how their personal information may be used, stored and accessed. Collecting and processing personal data from social media, by unmanned aerial vehicles (UAV), or public records (often under the \u0026ldquo;big data\u0026rdquo; category) that lack explicit consent are particularly problematic. Although the protection of privacy can be understood an essential right to safeguard human dignity [\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e], more studies and initiatives in humanitarian assistance need to resolve the apparent conflict between the duty to protect privacy and the urgent duty to assist and protect those in danger [\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eMany studies pointed out that organizations frequently collect much more data than they need (n\u0026thinsp;=\u0026thinsp;17) or are able to absorb (n\u0026thinsp;=\u0026thinsp;40). We consider the former as a potential for harm, as any excessive information increases the risks associated with data leaks and privacy violations. Data collected but that is not used can be regarded as related to the ethical value of beneficence, which implies that all information collected should have a concrete purpose related to informing humanitarian assistance. But even for data that were used for the intended purpose, a majority (n\u0026thinsp;=\u0026thinsp;61) of studies discussed that it may be too unreliable or inaccurate to adequately inform assistance programming.\u003c/p\u003e \u003cp\u003eA strong theme emerged regarding insufficient consent mechanisms, which strongly relates to the ethical value category of autonomy. About half the studies (n\u0026thinsp;=\u0026thinsp;52) mentioned that informed consent was either not provided by the affected population or was given without a full picture of how data would be processed or used. Eight studies cited that data might be used for reasons other than the original purpose for which consent may have been obtained. Related to the ethical value category of autonomy, about one in four papers (n\u0026thinsp;=\u0026thinsp;26) mentioned that a refusal to provide information could lead to being excluded from receiving assistance. This issue is illustrated by Shoemaker et al. [\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e] who documented how refugees felt that they lacked a choice on whether or not to provide personal information to UNHCR as their ability to access assistance depended on it. Detailed guidance has been created by the International Rescue Committee on how to obtain proper consent, whereas the International Committee of the Red Cross has published the legal basis for situations when data processing is permissible\u0026mdash;even when consent cannot be assumed or obtained [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e]. However, more work is clearly needed to train humanitarian professionals in these practices, and to monitor for better compliance with best consent practices as well as other minimal ethical guidelines. Existing guidance also needs to be updated to ensure the protection of private, personal and demographically identifiable information that extends to population groups rather than individuals [\u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eDirectly related to the value category of justice, a majority (n\u0026thinsp;=\u0026thinsp;63) of studies were concerned about data being processed in a way that may result in aid being distributed disproportionately with regard to people\u0026rsquo;s actual needs. This finding directly mirrors the importance of the humanitarian impartiality value category which refers to providing assistance solely based on need and regardless of personal preferences or discriminatory factors [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eA cross-cutting issue was the potential of data processing to exacerbate power imbalances (mentioned by 34 studies), often due to an exclusion from data collection, given the unequal access to certain technologies (n\u0026thinsp;=\u0026thinsp;14). In many cases, data processing was found to diminish the perceived neutrality of humanitarian organizations (n\u0026thinsp;=\u0026thinsp;29) as data could be processed in a way that might benefit one side of the conflict over the other. Concerningly, about half (n\u0026thinsp;=\u0026thinsp;48) of the studies found that humanitarian data processing might be overly dependent on potentially biased external entities (such as commercial entities, militaries, or foreign governments). This could be increasingly problematic for humanitarian organizations for multiple operational and ethical reasons, but particularly in conflict environments where the perception of independence is widely considered to be an essential humanitarian value category.\u003c/p\u003e \u003cp\u003eAnother theme identified across many studies was that data processing did not follow the principles of Accountability to Affected People (AAP) [\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e, \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e], which manifested in various ways across several of the four bioethical value categories. For the value category of autonomy, 18 studies remarked that affected communities were not involved in decisions on whether to use experimental technologies, whereas a smaller number (n\u0026thinsp;=\u0026thinsp;8) commented that processed information was not being made available to communities to allow for better group agency. Related to the values category of non-maleficence, nine papers discussed people\u0026rsquo;s inability to rectify inaccurate information about them or receive any form of compensation. Finally, related to the value category of justice, one in six (n\u0026thinsp;=\u0026thinsp;16) of studies found that data processing lacked accountability in terms of humanitarian organizations\u0026rsquo; obligation to protect rights\u0026mdash;or even pointed to ways that they may be violating these rights themselves. The Signal Code [\u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e], first published by the Harvard Humanitarian Initiative in 2017, considers data agency and redress/rectification as crucial rights and proposes specific actions to safeguard them in practice. We propose extending this list to always include affected communities when sharing collected data and involving them in decisions over experimental technologies.\u003c/p\u003e \u003cp\u003eAbout half of studies (n\u0026thinsp;=\u0026thinsp;52) cited ethical issues that were rooted in real-life experiences whereas almost one third (n\u0026thinsp;=\u0026thinsp;29) contained issues based on qualitative interviews or expert consultations. This signals that ethical issues have moved from theoretical concerns to actual incidents. However, it also reflects the large and diverse array of ethical issues that are emerging in connection with data processing in humanitarian crises which may first manifest as theoretical concerns before being validated as potentially negative consequences that can and do occur in real life.\u003c/p\u003e \u003cp\u003eThe results from this study show a wide array of ethical issues that should be addressed when processing data in a humanitarian context. However, to our knowledge, to date no humanitarian data protection guidance is sufficiently comprehensive to provide practical guidance for all concerns identified in the literature. More work may therefore be needed to expand existing guidelines and to ensure that future updates are also informed by systematic reviews of newly published studies. Moreover, there is a need to improve humanitarian organizations\u0026rsquo; accountability to affected populations, including the need to proactively prevent harm, and monitor the potential for causing harm and to limit risks. Finally, organizations involved in providing humanitarian assistance should review internal processes for training staff and create methods for verification that ensure appropriate minimal ethical standards are being met.\u003c/p\u003e \u003cp\u003eGeographically, publications were disproportionately from authors in high-income countries, primarily in Europe and North America, demonstrating a high level of interest in countries that have been the traditional source of most humanitarian funding but also of most technological innovation. Conversely, the small number of authors from lower-middle income countries and the complete absence of any authors from low-income countries highlights the lack of published perspectives from countries most affected by humanitarian crises. The lack of any study with an author from China may reflect that the large body of disaster related studies from Chinese authors published in English primarily discuss the response to domestic rather than foreign crises, or that ethical issues explored in this review may be more explored in Chinese language publications. People living in affected countries make up the vast majority of humanitarian organizations\u0026rsquo; staff, which could be a potential boon to a more diverse authorship on this subject. However, given the very small number of studies with authors from a humanitarian organization, more efforts need to be made by publishers to invite and support submissions from humanitarian professionals.\u003c/p\u003e \u003cp\u003eFrom 2010 to 2019, the number of peer-reviewed articles and conference proceedings about the ethical issues related to processing data of people affected by humanitarian crises has grown significantly, particularly after 2015. Published research shows widespread concerns that data processing in humanitarian assistance, without due ethical attention to known harms or potential risks, can cause additional harm, may not provide direct benefit, can show a lack of respect for affected populations\u0026rsquo; autonomy, and can lead to the unfair distribution of resources, among other concerns. We found that studies containing ethical discussions are often skewed towards investigating natural disaster contexts, as well as the use of technologies that allow the involvement of non-traditional actors, especially by gathering information from social media or crowdsourcing platforms. More research into ethical issues that arise in conflict settings is needed to better investigate the heightened security risks to vulnerable people in the large number of humanitarian crises associated with conflict, war, and social instability.\u003c/p\u003e \u003cp\u003eAs some studies in this review have pointed out, data may be collected in many humanitarian crises without due attention to data privacy and the security of the person providing the data, leaving them subject to malfeasant actors. Further, data can be collected or processed in a way that is inaccurate or out-of-date for the purpose for which it is being used, a risk that increases with methods such as crowdsourcing or analyzing social media posts. Both potential outcomes may pose added safety and security risks for people already vulnerable in a humanitarian crisis. They may also potentially render inaccurate needs assessments to humanitarian actors that may mean material humanitarian assistance delivered is inappropriate to actual needs, thus further increasing the vulnerability of people already at risk.\u003c/p\u003e \u003cp\u003eThe ethical issues identified in this review should be used to inform the development of ethical codes of conduct (whether voluntary or mandated by organizations). Further, companies and institutions behind the various technologies\u0026mdash;as well as the humanitarian organizations that use them to process data as part of their work\u0026mdash;should investigate to what extent these ethical issues are being addressed, and where more needs to be done. However, existing humanitarian data protection guidance and mechanisms are not sufficient to address all concerns identified in the literature and in this study. Likewise, training and accountability mechanisms to monitor the actual harm or potential for causing harm and to limit risks, are insufficient. These guidelines and mechanisms will need to be reviewed, expanded and informed by regular reviews that keep pace with technological change and changes in practice. Further research, especially using empirical methods, is necessary to better identify and understand the type and prevalence of ethical issues in the field.\u003c/p\u003e \u003cp\u003eFinally, more investigations are needed into the appropriate and inappropriate use of commonplace humanitarian tools and data management processes, such as CAPI, spreadsheets, filesharing, or use of online databases. At the same time, case studies of early adoptions of AI should address which ethical considerations were given when using tools that may involve data processing using multiple services and companies globally, in order to inform local decisions. Such research is urgently needed to create better guidance, training, and auditing methods to support humanitarian organizations to use data processing technologies as ethically as possible.\u003c/p\u003e \u003cp\u003eThe number of studies discussing natural disasters (n\u0026thinsp;=\u0026thinsp;37) was about the same as the number discussing armed conflict (n\u0026thinsp;=\u0026thinsp;36), even though by the end of 2020, 87% of displacement was caused by conflict [\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e]. This disproportionate focus may be due to disasters generating a higher level of media attention, as well as interest among technology enthusiasts, volunteers, and private companies\u0026mdash;a trend identified by several studies [\u003cspan additionalcitationids=\"CR80 CR81\" citationid=\"CR79\" class=\"CitationRef\"\u003e79\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e]. Likewise, empirical research in conflict settings is far more difficult given the inherent security risks, which in turn limits the development of theories and academic discourse that rely on data from the field. More research will be needed in the future focusing on ethical issues that are unique to conflict settings, as data processing without appropriate consideration of ethical issues in these settings arguably has the potential to cause far greater harm.\u003c/p\u003e \u003cp\u003eOur results show a significant focus on both internally and externally displaced populations, particularly those trying to reach Europe or the United States. Likewise, the Ebola virus disease outbreak in West Africa in 2014\u0026ndash;2016 was the focus of major international containment response because of the perceived threat of pandemic spread beyond the region, and the subject of a large number of well-funded studies. The increased focus on assistance to displaced populations could be due to the intense media coverage, whereas technological experimentation during the Ebola response was taken to new levels in areas such as processing of call data records without explicit consent [\u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e83\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eA significant number of studies discussed ethical issues without going into detail about the particular context: 20 studies discussed humanitarian assistance or crisis response in general, while 23 did not specify a data processing purpose. Given the lack of widely shared understandings of what constitutes terms such as \u0026ldquo;humanitarian community\u0026rdquo; or \u0026ldquo;information and communication technologies\u0026rdquo;, we recommend that even theoretical papers provide sufficient definitions and examples.\u003c/p\u003e \u003cp\u003eWe found that studies most commonly discussed activities involving the initial collection of data from affected populations, including assessments, registrations, and health interventions. To some extent, this reflects that a large number of studies investigated the use of crowdsourcing and social media to gain an understanding of a particular humanitarian crisis (see below). It may also be a reflection of the increasing emphasis that humanitarian organizations and their donors have been placing in recent years on establishing an \u0026ldquo;evidence base\u0026rdquo; before rolling out assistance programs [\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e, \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e85\u003c/span\u003e]. More research is needed to investigate the link between the potential increase of ethical risk and the push for collecting more needs assessment data.\u003c/p\u003e \u003cp\u003eStudies discussing social media (n\u0026thinsp;=\u0026thinsp;48), crowdsourcing (n\u0026thinsp;=\u0026thinsp;45), and mapping (n\u0026thinsp;=\u0026thinsp;43) dominated, often due to the perceived lack of good ground-validated data in humanitarian assistance. There were many use cases of social media, but the most-discussed application was mining public Twitter posts for clues on potential population needs. We also found that many studies focus on the potential use of other \u0026ldquo;new\u0026rdquo; technologies, especially if they can be used remotely to assess needs (e.g., satellite imagery, unmanned aerial vehicles, call data records). Crowdsourcing, a method of obtaining information from the general public [\u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e86\u003c/span\u003e], was discussed by almost half the studies. Many studies traced their enthusiasm for\u0026mdash;or criticism of\u0026mdash;crowdsourcing to the creation of the Ushahidi platform (mentioned by 25 studies) in 2007. Similarly, the emergence of digital platform based volunteer networks since the 2010 Haiti earthquake [\u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e87\u003c/span\u003e, \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e88\u003c/span\u003e] can partially explain the large number of studies referencing these tools.\u003c/p\u003e \u003cp\u003eSurprisingly, only three studies mentioned computer-assisted personal interviewing (CAPI) tools such as KoboToolbox which has been adopted by a broad range of international and national humanitarian agencies as the tool of choice for needs assessments [\u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e, \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e90\u003c/span\u003e]. Similarly, use of spreadsheets were only mentioned by one study as a cause for ethical concern, despite being the main data storage and sharing mechanism of choice for many humanitarian organizations [\u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e91\u003c/span\u003e]. Such low-tech data processing means are addressed in recent guidelines, for example, giving guidance on how to remove sensitive data before sharing Excel files with others [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. However, more research is needed on current practices and ethical risks associated with these commonly used technologies.\u003c/p\u003e \u003cp\u003eThe ethical issues associated with biometrics were discussed by a significant number of studies, particularly for the registration of refugees and other migrants by organizations such as the UN High Commissioner for Refugees, UNHCR [see, for example, 13]. In 2015, such concerns led Oxfam, one of the largest international humanitarian NGOs, to put a moratorium on its use of biometrics in order to assess potential risks [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. In 2021, this in turn resulted in the creation of a policy intended to ensure that the technology is used ethically within Oxfam\u0026rsquo;s operations [\u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e92\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eFinally, ethics related to AI and similar technologies were discussed significantly more frequently after 2017. This seems to correlate with the growing presence and desire to analyze \u0026ldquo;big data\u0026rdquo; resources, in order to learn more about the needs and sentiments of the affected population. For example, big data and medical data were mentioned twice as frequently by studies that discussed AI; call data records were cited 78% more often. More theoretical and empirical research is needed to understand the potential issues that come with applying these rapidly evolving technologies to humanitarian assistance.\u003c/p\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003eLimitations of the Scoping Review\u003c/h2\u003e \u003cp\u003eThis study is limited to literature published since 2010 and before January 2020, and it excludes work from non-peer-reviewed sources. As mentioned above, identifying all relevant studies was a significant challenge due to the lack of a shared nomenclature across disciplines for humanitarian assistance, ethical issues, and data processing. As a result, potentially relevant articles that met the inclusion criteria may have been missed. Nonetheless, we believe that our search strategy represents the most comprehensive and inclusive set of keywords to capture studies in the diverse field of humanitarian assistance to date.\u003c/p\u003e \u003cp\u003eAs suggested by the Arksey and O\u0026rsquo;Malley framework, a consultation exercise with humanitarian and ethics experts will be organized to present our results, aid knowledge translation, ensure that the results from this study are relevant, and frame a future research agenda. The results of this consultation will be published separately.\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThis extensive review of the literature highlights a growing concern over ethical challenges in data processing within humanitarian contexts, including those related to the increasing use of AI. Our findings underscore significant ethical risks associated with data processing in these settings, including potential harm, lack of direct benefits, infringement on populations\u0026rsquo; autonomy, and the unfair allocation of resources. Notably, a quarter of the studies reviewed address AI technologies, pointing to privacy concerns as the most frequent ethical issue, especially regarding the inadvertent sharing of sensitive data with third parties.\u003c/p\u003e \u003cp\u003eThe underrepresentation of perspectives from low and middle-income countries in the academic discourse further exacerbates these challenges, highlighting the urgent need for more diverse and inclusive perspectives. Additional research, especially using empirical methods, is necessary to better identify and understand the type and prevalence of ethical issues in the field. While natural disasters predominate the literature, more studies are needed investigate the unique ethical issues that arise in conflict settings to better address the heightened security risks to vulnerable people in war.\u003c/p\u003e \u003cp\u003eExisting humanitarian data protection guidance as well as training and accountability methods for monitoring potential harm and to limit risks are insufficient to address all concerns identified in the literature and in this study. These guidelines and mechanisms will need to be reviewed, expanded, and informed by regular reviews that keep pace with technological change and changes in practice. Likewise, companies and institutions behind the various technologies\u0026mdash;as well as the humanitarian organizations that use them to process data as part of their work\u0026mdash;should investigate to what extent the ethical issues identified in this study are being addressed, and where more needs to be done.\u003c/p\u003e \u003cp\u003eFinally, investigations are urgently needed into early adoptions of AI tools in humanitarian contexts, including the rapid spread of large language models such as ChatGPT, to ensure these technologies are harnessed with utmost ethical rigor, safeguarding the dignity and rights of those in crisis while enhancing the efficacy and fairness of humanitarian responses.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"625\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.08%\"\u003e\n \u003cp\u003e\u003cstrong\u003eAcronym\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"68%\"\u003e\n \u003cp\u003e\u003cstrong\u003eDefinition\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.92%\"\u003e\n \u003cp\u003e\u003cstrong\u003ePage\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.08%\"\u003e\n \u003cp\u003eAAP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"68%\"\u003e\n \u003cp\u003eAccountability to Affected People\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.92%\"\u003e\n \u003cp\u003e23\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.08%\"\u003e\n \u003cp\u003eCAPI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"68%\"\u003e\n \u003cp\u003eComputer-assisted personal interviewing\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.92%\"\u003e\n \u003cp\u003e26\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.08%\"\u003e\n \u003cp\u003eCOVID-19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"68%\"\u003e\n \u003cp\u003eCoronavirus disease (2019)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.92%\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.08%\"\u003e\n \u003cp\u003eGDPR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"68%\"\u003e\n \u003cp\u003eGeneral Data Protection Regulation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.92%\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.08%\"\u003e\n \u003cp\u003eGIS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"68%\"\u003e\n \u003cp\u003eGeographic information system\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.92%\"\u003e\n \u003cp\u003e16\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.08%\"\u003e\n \u003cp\u003eICT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"68%\"\u003e\n \u003cp\u003eInformation and communication technology\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.92%\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.08%\"\u003e\n \u003cp\u003eOCHA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"68%\"\u003e\n \u003cp\u003eUnited Nations Office for the Coordination of Humanitarian Affairs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.92%\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.08%\"\u003e\n \u003cp\u003ePRISMA-ScR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"68%\"\u003e\n \u003cp\u003efollowed Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.92%\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.08%\"\u003e\n \u003cp\u003eSMS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"68%\"\u003e\n \u003cp\u003eShort messaging service\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.92%\"\u003e\n \u003cp\u003e17\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.08%\"\u003e\n \u003cp\u003eUAV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"68%\"\u003e\n \u003cp\u003eUnmanned aerial vehicles\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.92%\"\u003e\n \u003cp\u003e17\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.08%\"\u003e\n \u003cp\u003eUNHCR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"68%\"\u003e\n \u003cp\u003eUnited Nations High Commissioner for Refugees\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.92%\"\u003e\n \u003cp\u003e23\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003e\u003cem\u003eEthics Approval and Consent to Participate\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003eConsent for Publication\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003eAvailability of Data and Materials\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data generated or analyzed during this study are included in this published article and its supplementary information files.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003eCompeting Interests\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003eFunding\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFunding from Grand Challenges Canada provided support for the second screening reviewer\u0026rsquo;s costs during the scoping review. The funder had no involvement in the study\u0026rsquo;s design, data collection, or analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003eAuthors\u0026apos; Contributions\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTK led the design of the study, conducted the data collection and analysis, and drafted the initial manuscript. JO provided overall supervision and mentorship to ensure the quality and integrity of the research. LA, PV, and JO provided guidance on the scoping review methodology and assessed the study protocol critically. JO and PV contributed to the definitions of ethical value categories and related concepts. AA contributed to the conceptualization of the study in relation to artificial intelligence and helped refine the research questions. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003eAcknowledgements\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003eAuthors\u0026apos; Information (optional)\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eOCHA. Global Humanitarian Overview 2023 [Internet]. Office for the Coordination of Humanitarian Affairs; 2022. Available from: https://reliefweb.int/report/world/global-humanitarian-overview-2023-enaresfr.\u003c/li\u003e\n\u003cli\u003eOCHA. Financial tracking service. 2023. https://fts.unocha.org. Accessed 2 Jul 2023.\u003c/li\u003e\n\u003cli\u003eHumanitarian Outcomes. Aid worker security database. 2021. https://aidworkersecurity.org/incidents/report. Accessed 1 Sep 2021.\u003c/li\u003e\n\u003cli\u003eDonini A, Maxwell D. From face-to-face to face-to-screen: remote management, effectiveness and accountability of humanitarian action in insecure environments. Int Rev Red Cross. 2013;95:383-413.\u003c/li\u003e\n\u003cli\u003eSchofield G, Dittborn M, Selman LE, Huxtable R. Defining ethical challenge(s) in healthcare research: a rapid review. BMC Med Ethics. 2021;22:135.\u003c/li\u003e\n\u003cli\u003eUN OCHA. Glossary of humanitarian terms. Geneva/New York: OCHA; 2003.\u003c/li\u003e\n\u003cli\u003eIASC. Protection of internally displaced persons. Geneva: IASC; 1999.\u003c/li\u003e\n\u003cli\u003eWHO. Risk reduction and emergency preparedness. Geneva: World Health Organization; 2007.\u003c/li\u003e\n\u003cli\u003eOCHA. Data responsibility guidelines. Geneva: UN Office for the Coordination of Humanitarian Affairs; 2019.\u003c/li\u003e\n\u003cli\u003eSlim H. Humanitarian ethics: a guide to the morality of aid in war and disaster. New York, NY: Oxford University Press; 2015.\u003c/li\u003e\n\u003cli\u003eInternational Committee of the Red Cross and Privacy International. The Humanitarian Metadata Problem: \u0026ldquo;Doing No Harm\u0026rdquo; in the Digital Era [Internet]. Geneva; 2018. Available from: https://privacyinternational.org/sites/default/files/2018-12/The Humanitarian Metadata Problem - Doing No Harm in the Digital Era.pdf.\u003c/li\u003e\n\u003cli\u003eSandvik KB, Gabrielsen Jumbert M, Karlsrud J, Kaufmann M. Humanitarian technology: a critical research agenda. Int Rev Red Cross. 2014;96:219-42.\u003c/li\u003e\n\u003cli\u003eJacobsen KL. Experimentation in humanitarian locations: UNHCR and biometric registration of Afghan refugees. Secur Dialogue. 2015;46:144-64.\u003c/li\u003e\n\u003cli\u003ePerakslis ED. Using digital health to enable ethical health research in conflict and other humanitarian settings. Confl Health. 2018;12:23.\u003c/li\u003e\n\u003cli\u003evan Wynsberghe A, Comes T. Drones in humanitarian contexts, robot ethics, and the human\u0026ndash;robot interaction. Ethics Inf Technol. 2019;22:43-53.\u003c/li\u003e\n\u003cli\u003eUN Global Pulse. Big Data for Development and Humanitarian Action: Towards Responsible Governance: Global Pulse Privacy Advisory Group Meetings [Internet]. New York, NY; 2016. Available from: http://unglobalpulse.org/sites/default/files/Big_Data_for_Development_and_Humanitarian_Action_Report_Final_0.pdf.\u003c/li\u003e\n\u003cli\u003eRaymond NA, Al Achkar Z, Verhulst S, Berens J. Building data responsibility into humanitarian action. OCHA Policy Stud Ser [Internet]. 2016;18. Available from: https://docs.unocha.org/sites/dms/Documents/TB18_Data Responsibility_Online.pdf. \u003c/li\u003e\n\u003cli\u003eThe Engine Room and Oxfam. Biometrics in the Humanitarian Sector [Internet]. 2018. Available from: https://www.theengineroom.org/wp-content/uploads/2018/03/Engine-Room-Oxfam-Biometrics-Review.pdf. \u003c/li\u003e\n\u003cli\u003eParker B. New UN deal with data mining firm Palantir raises protection concerns. 2019. https://www.irinnews.org/news/2019/02/05/un-palantir-deal-data-mining-protection-concerns-wfp. Accessed 25 Aug 2021.\u003c/li\u003e\n\u003cli\u003eIASC. Operational guidance: data responsibility in humanitarian action. Geneva: Inter-Agency Standing Committee; 2021.\u003c/li\u003e\n\u003cli\u003eInternational Committee of the Red Cross. Handbook on Data Protection in Humanitarian Action (2nd Edition) [Internet]. 2020. Available from: https://www.icrc.org/en/data-protection-humanitarian-action-handbook.\u003c/li\u003e\n\u003cli\u003eKrishnaraj G, Hunt M, Schwartz L, O\u0026rsquo;Math\u0026uacute;na D, Pringle J, Okhowat A, et al. Ethics for humanitarian innovation: background paper. London: Elrha; 2021.\u003c/li\u003e\n\u003cli\u003eElhra, Humanitarian Health Ethics. Ethics for humanitarian innovation: Toolkit [Internet]. London; 2021. Available from: https://higuide.elrha.org/wp-content/uploads/2021/09/Elrha-Ethics_for_Humanitarian_Innovation-Toolkit-210913.pdf.\u003c/li\u003e\n\u003cli\u003eTegmark M. Life 3.0 : being human in the age of artificial intelligence. New York, NY: Alfred A. Knopf; 2017.\u003c/li\u003e\n\u003cli\u003eWong F, Zheng EJ, Valeri JA, Donghia NM, Anahtar MN, Omori S, et al. Discovery of a structural class of antibiotics with explainable deep learning. Nature. 2023. doi: 10.1038/s41586-023-06887-8.\u003c/li\u003e\n\u003cli\u003eOpenAI. ChatGPT [Internet]. 2024. Available from: https://chat.openai.com/.\u003c/li\u003e\n\u003cli\u003eStahl BC, Eke D. The ethics of ChatGPT \u0026ndash; Exploring the ethical issues of an emerging technology. Int J Inf Manag. 2024;74:102700.\u003c/li\u003e\n\u003cli\u003eDave T, Athaluri SA, Singh S. ChatGPT in medicine: an overview of its applications, advantages, limitations, future prospects, and ethical considerations. Front Artif Intell. 2023;6:1169595.\u003c/li\u003e\n\u003cli\u003eShen Y, Heacock L, Elias J, Hentel KD, Reig B, Shih G, et al. ChatGPT and other large language models are double-edged swords. Radiology. 2023;307:e230163.\u003c/li\u003e\n\u003cli\u003eTansey CM, Pringle J, Dav\u0026eacute; A, Boulanger R, Hunt M. Earthquakes to floods: a scoping review of health-related disaster research in low- and middle-income countries. PLoS Curr. 2018;10. doi: 10.1371/currents.dis.57d98a902a326361d88d54521e68b016.\u003c/li\u003e\n\u003cli\u003eMakhoul J, Chehab RF, Shaito Z, Sibai AM. A scoping review of reporting \u0026lsquo;Ethical Research Practices\u0026rsquo; in research conducted among refugees and war-affected populations in the Arab world. BMC Medical Ethics. 2018;19:36.\u003c/li\u003e\n\u003cli\u003ePal NE, Eckenwiler L, Hyppolite SR, Pringle J, Chung R, Hunt M. Ethical considerations for closing humanitarian projects: a scoping review. J Int Humanit Action. 2019;4:17.\u003c/li\u003e\n\u003cli\u003eWatson H, Rodrigues R. Bringing privacy into the fold: considerations for the use of social media in crisis management. J Contingencies Crisis Manag. 2017;26:89-98.\u003c/li\u003e\n\u003cli\u003eBempong NE, Casta\u0026ntilde;eda RRD, Sch\u0026uuml;tte S, Bolon I, Keiser O, Escher G, et al. Precision global health - the case of ebola: a scoping review. J Glob Health. 2019;9:010404.\u003c/li\u003e\n\u003cli\u003eMesmar S, Talhouk R, Akik C, Olivier P, Elhajj IH, Elbassuoni S, et al. The impact of digital technology on health of populations affected by humanitarian crises: recent innovations and current gaps. J Public Health Policy. 2016;37:167-200.\u003c/li\u003e\n\u003cli\u003eWang N, Christen M, Hunt M. Ethical considerations associated with \u0026ldquo;humanitarian drones\u0026rdquo;: a scoping literature review. Sci Eng Ethics. 2021;27:51.\u003c/li\u003e\n\u003cli\u003ePictet J. The fundamental principles of the red cross: commentary. Geneva: ICRC; 1979.\u003c/li\u003e\n\u003cli\u003eMSF. Independent, impartial, neutral [Internet]. [cited 2021 Jan 4]. Available from: https://www.doctorswithoutborders.org/who-we-are/principles/independence. \u003c/li\u003e\n\u003cli\u003eOCHA. Humanitarian principles. New York: UN Office for the Coordination of Humanitarian Affairs; 2012.\u003c/li\u003e\n\u003cli\u003eIFRC. The code of conduct for the international red cross and red crescent movement and non governmental organisations (NGOs) in disaster relief. Geneva: IFRC; 1994.\u003c/li\u003e\n\u003cli\u003eSphere Association. The sphere handbook: humanitarian charter and minimum standards in humanitarian response. Geneva: Practical Action Publishing; 2018.\u003c/li\u003e\n\u003cli\u003eHilhorst D, Schmiemann N. Humanitarian principles and organisational culture: everyday practice in Me\u0026eacute;decins Sans Fronti\u0026eacute; res-Holland. Dev Pract. 2002;12:490-500.\u003c/li\u003e\n\u003cli\u003eRaymond N, Card B. Applying humanitarian principles to current uses of information communication technologies: gaps in doctrine and challenges to practice. Cambridge, MA: Harvard Humanitarian Initiative; 2015.\u003c/li\u003e\n\u003cli\u003eBroussard G, Rubenstein LS, Robinson C, Maziak W, Gilbert SZ, DeCamp M. Challenges to ethical obligations and humanitarian principles in conflict settings: a systematic review. J Int Humanit Action. 2019;4:15.\u003c/li\u003e\n\u003cli\u003eBeauchamp T, Childress J. Principles of biomedical ethics. New York, NY: Oxford University Press; 2019.\u003c/li\u003e\n\u003cli\u003eTrials of war criminals before the Nuernberg military tribunals under Control Council law no. 10 (Volume 2) [Internet]. Washington, D.C.: U.S. Government Printing Office; 1949. Available from: https://collections.wip.nlm.nih.gov/catalog/nlm:nlmuid-01130400RX2-mvpart. \u003c/li\u003e\n\u003cli\u003eWorld Medical Association. World medical association declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013;310:2191-4.\u003c/li\u003e\n\u003cli\u003eNational Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. Belmont report: ethical principles and guidelines for the protection of human subjects of research. Washingtion, D.C.: Department of Health, Education, and Welfare; 1978.\u003c/li\u003e\n\u003cli\u003ePham P, Vinck P. Technology fusion and their implications for conflict early warning systems, public health, and human rights. Health Hum Rights. 2012;14:106-17.\u003c/li\u003e\n\u003cli\u003eCawthorne D, Wynsberghe ARV. From HealthDrone to FrugalDrone: value-sensitive design of a blood sample transportation drone. In: 2019 IEEE international symposium on technology and society (ISTAS). Medford, MA, USA: IEEE; 2019. p. 1-7.\u003c/li\u003e\n\u003cli\u003ePeters MDJ, Godfrey CM, Khalil H, McInerney P, Parker D, Soares CB. Guidance for conducting systematic scoping reviews. Int J Evid Based Healthc. 2015;13:141-6.\u003c/li\u003e\n\u003cli\u003eArksey H, O\u0026apos;Malley L. Scoping studies: towards a methodological framework. Int J Soc Res Methodol. 2005;8:19-32.\u003c/li\u003e\n\u003cli\u003eLevac D, Colquhoun H, O\u0026apos;Brien KK. Scoping studies: advancing the methodology. Implement Sci. 2010;5:69.\u003c/li\u003e\n\u003cli\u003ePeters MDJ, Godfrey C, Mcinerney P, Baldini Soares C, Khalil H, Parker D. Chapter 11: Scoping reviews. In: Aromataris E, Munn Z, editors. Joana Briggs Inst Rev Man [Internet]. 2017. p. 6\u0026ndash;24. Available from: https://reviewersmanual.joannabriggs.org/ \u003c/li\u003e\n\u003cli\u003eTricco AC, Lillie E, Zarin W, O\u0026apos;Brien KK, Colquhoun H, Levac D, et al. PRISMA extension for scoping reviews (PRISMA-ScR): checklist and explanation. Ann Intern Med. 2018;169:467-73.\u003c/li\u003e\n\u003cli\u003eWorld Bank. World bank country and lending groups. Washingtion, D.C.: The World Bank; 2020.\u003c/li\u003e\n\u003cli\u003eUN News Centre. Interview: senior UN relief official warns no room for complacency until Ebola is eliminated. 2015. https://ebolaresponse.un.org/interview-senior-un-relief-official-warns-no-room-complacency-until-ebola-eliminated. Accessed 1 Jun 2021.\u003c/li\u003e\n\u003cli\u003eOCHA. Financial Tracking Service [Internet]. 2023 [cited 2023 Jul 2]. Available from: https://fts.unocha.org. \u003c/li\u003e\n\u003cli\u003eEvidence Partners. DistillerSR [Internet]. Ottawa, Canada; 2022. Available from: https://www.evidencepartners.com. \u003c/li\u003e\n\u003cli\u003eShoemaker E, Kristinsdottir GS, Ahuja T, Baslan D, Pon B, Currion P, et al. Identity at the margins: examining refugee experiences with digital identity systems in Lebanon, Jordan, and Uganda. In: Proceedings of the 2nd ACM SIGCAS conference on computing and sustainable societies. New York, NY, USA: ACM; 2019. p. 206-17.\u003c/li\u003e\n\u003cli\u003ePaul S, Sosale S. Witnessing a disaster: public use of digital technologies in the 2015 South Indian floods. Digit J. 2020;8:15-31.\u003c/li\u003e\n\u003cli\u003eHayes P, Kelly S. Distributed morality, privacy, and social media in natural disaster response. Technol Soc. 2018;54:155-67.\u003c/li\u003e\n\u003cli\u003eLatif S, Qayyum A, Usama M, Qadir J, Zwitter A, Shahzad M. Caveat emptor: the risks of using big data for human development. IEEE Technol Soc Mag. 2019;38:82-90.\u003c/li\u003e\n\u003cli\u003eComes T, Meesters K, Torjesen S. Making sense of crises: the implications of information asymmetries for resilience and social justice in disaster-ridden communities. Sustain Resilient Infrastruct. 2019;4:124-36.\u003c/li\u003e\n\u003cli\u003eHaworth BT, Bruce E, Whittaker J, Read R. The good, the bad, and the uncertain: contributions of volunteered geographic information to community disaster resilience. Front Earth Sci. 2018;6:183.\u003c/li\u003e\n\u003cli\u003eVannini S, Gomez R, Lopez D, Mora S, Morrison JC, Tanner J, et al. Humanitarian organizations\u0026apos; information practices: procedures and privacy concerns for serving the undocumented. E J Info Sys Dev Countries. 2019;86:e12109.\u003c/li\u003e\n\u003cli\u003eCinnamon J, Jones SK, Adger WN. Evidence and future potential of mobile phone data for disease disaster management. Geoforum. 2016;75:253-64.\u003c/li\u003e\n\u003cli\u003eAnderson M. Do no harm: how aid can support peace \u0026ndash; Or war. Boulder, CO: Lynne Rienner Publishers; 1999.\u003c/li\u003e\n\u003cli\u003eWHS. The Grand Bargain \u0026ndash; a shared commitment to better serve people in need. Istanbul: World Humanitarian Summit; 2016.\u003c/li\u003e\n\u003cli\u003eSandvik K, Raymond N. Beyond the protective effect: towards a theory of harm for information communication technologies in mass atrocity response. Genocide Stud Prev. 2017;11:9-24.\u003c/li\u003e\n\u003cli\u003eUN General Assembly. Universal Declaration of Human Rights [Internet]. Gen. Assem. Resolut. 1948 [cited 2022 Feb 14]. Available from: https://www.un.org/en/ga/search/view_doc.asp?symbol=A/RES/217(III). \u003c/li\u003e\n\u003cli\u003eFloridi L. On human dignity as a foundation for the right to privacy. Philos Technol. 2016;29:307-12.\u003c/li\u003e\n\u003cli\u003eInternational Rescue Committee. Obtaining Meaningful Informed Consent [Internet]. New York; 2018. Available from: https://www.principlesinpractice.info/help-library/irc-research-toolkit-obtaining-meaningful-informed-consent. \u003c/li\u003e\n\u003cli\u003eRaymond NA. Beyond \u0026ldquo;do no harm\u0026rdquo; and individual consent: reckoning with the emerging ethical challenges of civil society\u0026rsquo;s use of data. In: Taylor L, Floridi L, VanDerSloot B, editors. Group privacy: new challenges of data technologies. Cham: Springer International Publishing; 2016. p. 67-82.\u003c/li\u003e\n\u003cli\u003eInternational Committee of the Red Cross. Accountability to Affected People Institutional Framework [Internet]. 2019. Available from: https://www.icrc.org/en/publication/accountability-affected-people-institutional-framework. \u003c/li\u003e\n\u003cli\u003eGlobal Health Cluster. Operational Guidance on Accountability to Affected Populations (AAP) [Internet]. 2017. Available from: https://www.humanitarianresponse.info/system/files/documents/files/cluster_coordination_reference_module_2015_final.pdf. \u003c/li\u003e\n\u003cli\u003eGreenwood F, Howarth C, Poole D, Raymond N, Scarnecchia D. The signal code: a human rights approach to information during crisis. Cambridge, MA: Harvard Humanitarian Initiative; 2017.\u003c/li\u003e\n\u003cli\u003eIDMC. 2021 global report on internal displacement. Geneva: Internal Displacement Monitoring Centre; 2021.\u003c/li\u003e\n\u003cli\u003eBurns R. New frontiers of philanthro‐capitalism: digital technologies and humanitarianism. Antipode. 2019;51:1101-22.\u003c/li\u003e\n\u003cli\u003eAretz YL. Data philanthropy. SSRN Electron J. 2019. doi: 10.2139/ssrn.3320798.\u003c/li\u003e\n\u003cli\u003eTaylor L. The ethics of big data as a public good: which public? Whose good? Philos Trans R Soc A Math Phys Eng Sci. 2016;374:20160126.\u003c/li\u003e\n\u003cli\u003eDuffield M. The resilience of the ruins: towards a critique of digital humanitarianism. Resilience. 2016;4:147-65.\u003c/li\u003e\n\u003cli\u003eMcDonald S. Ebola: A Big Data Disaster [Internet]. Bengaluru and Delhi; 2016. Report No.: 2016.01. Available from: https://cis-india.org/papers/ebola-a-big-data-disaster. \u003c/li\u003e\n\u003cli\u003eBlanchet K, Ramesh A, Frison S, Warren E, Hossain M, Smith J, et al. Evidence on public health interventions in humanitarian crises. Lancet. 2017;390:2287-96.\u003c/li\u003e\n\u003cli\u003ePham KT, Sattigeri P, Dhurandhar A, Jacob AC, Vukovic M, Chataigner P, et al. Real-time understanding of humanitarian crises via targeted information retrieval. IBM J Res Dev. 2017;61:7:1-7:12.\u003c/li\u003e\n\u003cli\u003eMartin-Shields C. The technologists dilemma: ethical challenges of using crowdsourcing technology in conflict and disaster-affected regions. Georget J Int Aff. 2013;14:157-63.\u003c/li\u003e\n\u003cli\u003ePhillips J. Risk in a digital age: understanding risk in virtual networks through digital response networks (DRNs). Int Dev Plan Rev. 2018;40:239-72.\u003c/li\u003e\n\u003cli\u003eMeier P. Digital humanitarians : how big data is changing the face of humanitarian response. Boca Raton: CRC Press; 2014.\u003c/li\u003e\n\u003cli\u003eOCHA. World humanitarian data and trends 2015. New York, NY: OCHA; 2015.\u003c/li\u003e\n\u003cli\u003eSapkota PP, Siddiqi K. Is ubiquitous technology for needs data management a game changer in humanitarian arena? Int J Inf Syst Crisis Response Manag. 2019;11:83-97.\u003c/li\u003e\n\u003cli\u003eMadon S, Schoemaker E. Reimagining refugee identity systems: a sociological approach. In: Nielsen P, Kimaro HC, editors. Information and communication technologies for development Strengthening southern-driven cooperation as a catalyst for ICT4D. Cham: Springer International Publishing; 2019. p. 660-74.\u003c/li\u003e\n\u003cli\u003eEaton-Lee J, Shaughnessy E. Oxfam\u0026rsquo;s new policy on biometrics explores safe and responsible data practice. 2021. https://reliefweb.int/report/world/oxfam-s-new-policy-biometrics-explores-safe-and-responsible-data-practice. Accessed 30 Jun 2021.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-ethics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"meth","sideBox":"Learn more about [BMC Medical Ethics](http://bmcmedethics.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/meth/default.aspx","title":"BMC Medical Ethics","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Humanitarian data, Ethical issues, Artificial intelligence (AI), Data processing, Ethical risks, Scoping review, Bioethical principles, Resource distribution","lastPublishedDoi":"10.21203/rs.3.rs-4224535/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4224535/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eHumanitarian organizations are rapidly expanding their use of data in the pursuit of operational gains in effectiveness and efficiency. Ethical risks, particularly from artificial intelligence (AI) data processing, are increasingly recognized yet inadequately addressed by current humanitarian data protection guidelines. This study reports on a scoping review that maps the range of ethical issues that have been raised in the academic literature regarding data processing of people affected by humanitarian crises.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eWe systematically searched databases to identify peer-reviewed studies published since 2010. Data and findings were standardized, grouping ethical issues into the value categories of autonomy, beneficence, non-maleficence, and justice. The study protocol followed Arksey and O\u0026rsquo;Malley\u0026rsquo;s approach and PRISMA reporting guidelines.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eWe identified 8,387 unique records and retained 98 relevant studies. One in four (n\u0026thinsp;=\u0026thinsp;25) discussed technologies related to artificial intelligence. Seven studies included an author from a lower-middle income country while none included an author from a low-income country. We identified 22 ethical issues which were then grouped along the four ethical value categories of autonomy, beneficence, non-maleficence, and justice. Slightly over half of included studies (n\u0026thinsp;=\u0026thinsp;52) identified ethical issues based on real-world examples. The most-cited ethical issue (n\u0026thinsp;=\u0026thinsp;74) was a concern for privacy in cases where personal or sensitive data might be inadvertently shared with third parties. The technologies most frequently discussed in these studies included social media, crowdsourcing, and mapping tools.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eStudies highlight significant concerns that data processing in humanitarian contexts can cause additional harm, may not provide direct benefits, may limit affected populations\u0026rsquo; autonomy, and can lead to the unfair distribution of scarce resources. The anticipated increase in AI tool deployment for humanitarian assistance amplifies these concerns. Urgent development of specific, comprehensive guidelines, training, and auditing methods are required to address these ethical challenges. Moreover, empirical research from low and middle-income countries, disproportionally affected by humanitarian crises, is vital to ensure inclusive and diverse perspectives. This research should focus on the ethical implications of both emerging AI systems as well as established humanitarian data management practices.\u003c/p\u003e\u003ch2\u003eTrial registration:\u003c/h2\u003e \u003cp\u003eNot applicable.\u003c/p\u003e","manuscriptTitle":"Ethical implications related to processing of personal data and artificial intelligence in humanitarian crises: A scoping review","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-04-19 17:22:46","doi":"10.21203/rs.3.rs-4224535/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-07-11T04:40:45+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-07-10T11:39:35+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"24585876565381424722357986850559747939","date":"2024-07-02T14:42:59+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-06-19T01:53:01+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"223901522032008940840014757417559617422","date":"2024-06-13T10:01:38+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-06-13T09:31:26+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2024-04-10T03:36:19+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-04-10T03:33:32+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-04-10T03:33:32+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Medical Ethics","date":"2024-04-05T18:49:05+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-ethics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"meth","sideBox":"Learn more about [BMC Medical Ethics](http://bmcmedethics.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/meth/default.aspx","title":"BMC Medical Ethics","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a406ba21-c4e6-4477-936b-26fc2a6e5e6e","owner":[],"postedDate":"April 19th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-04-21T15:58:45+00:00","versionOfRecord":{"articleIdentity":"rs-4224535","link":"https://doi.org/10.1186/s12910-025-01189-2","journal":{"identity":"bmc-medical-ethics","isVorOnly":false,"title":"BMC Medical Ethics"},"publishedOn":"2025-04-15 15:57:01","publishedOnDateReadable":"April 15th, 2025"},"versionCreatedAt":"2024-04-19 17:22:46","video":"","vorDoi":"10.1186/s12910-025-01189-2","vorDoiUrl":"https://doi.org/10.1186/s12910-025-01189-2","workflowStages":[]},"version":"v1","identity":"rs-4224535","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4224535","identity":"rs-4224535","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}