Generative AI’s Impact on Organizational Structures: An Analysis in Collaboration with ChatGPT

Generative AI’s Impact on Organizational Structures: An Analysis in Collaboration with ChatGPT | 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 Generative AI’s Impact on Organizational Structures: An Analysis in Collaboration with ChatGPT Chloe Latto, Alexander Richter, Mary Tate, Michael Leyer This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6649663/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Artificial intelligence has been rapidly increasing in sophistication and impact, reinforcing discussions of its extensive potential to disrupt jobs and organizations. This paper investigates the disruptive potential of generative artificial intelligence (genAI) in organizations by investigating its impact on organizational structures for core business functions as work re-design occurs with the introduction of genAI. Our analysis investigates the impacts of genAI, but we also use genAI as a partner, in collaboration with the authors. Data from the O*Net 2023 database is utilized to consolidate a list of 85 corporate knowledge worker roles and (associated skill levels), as part of our analysis, ChatGPT contributed its assessment as to which roles genAI will take over partly or fully in the near future. Our analysis led to a “flat” organization structure which allowed the surfacing of the extent of impact genAI will have on roles and, in turn, organizational structures. Results suggest that use of genAI is accelerating the trend of flattening hierarchies and will lead to more independent teams with joint Human-AI capability. Finally, we observe that AI is already on a disruptive trajectory. Individuals and organizations not engaging with the potential of genAI, and the expected organizational changes, are at serious risk of underperformance, role redundancy, and negative outcomes from the uncontrolled use of genAI. Generative Artificial Intelligence Organizational Structures ChatGPT Re-designing Roles Figures Figure 1 Figure 2 Figure 3 Figure 4 1 Introduction Generative artificial intelligence (GenAI), is rapidly transforming organizational landscapes. Unlike earlier waves of AI and digitalization, which primarily targeted routine tasks, genAI is now poised to revolutionize areas previously untouched by automation, including those requiring basic creativity and nuanced communication (Ghatak, 2023). These advancements are not only reshaping individual roles but are also expected to drive significant changes in organizational structures. Organizational structures ensure that responsibilities and functional areas are efficiently assigned to employees since an individual at the top of an organization (typically the CEO) cannot oversee everything in person. Thus, it is shared. Moreover, humans can only actively manage a certain number of social contacts, and personally interact with a limited number, so organizational units are required to ensure functionality of an organization. The introduction of genAI has enabled two main aspects in organizations. First, knowledge in the organization can be made easily accessible to every member of an organization, even if one has just started the job. Second, repetitive tasks can be completed in part by machines. As a result, more activities can be completed by genAI, with activities requiring human judgement remaining for humans (Hinsliff, 2023). However, to date, there has been no holistic analysis of the impacts of genAI on organization structures for core organizational functions, (Latto, et al., 2025), based on an analysis of its impacts on specific roles. Hence, in our study, we inquire how knowledge worker roles will change, re-designing work and how these changes will affect organizational structures. Utilizing AI tools is not just a current issue for knowledge work organizations; it is also becoming increasingly relevant in academic research. Tools such as ChatGPT, NotebookLM, and Microsoft’s Copilot provide new opportunities, but their emergence also raises important questions about the research process and ethics. While studies, including those summarized by Rahman et al. (2023), indicate that genAI is not yet suitable for conducting full empirical studies, they argue it can be effectively utilized for specific elements of the research process. In this paper, we explore this idea by integrating ChatGPT as a member of the research team alongside four human researchers to assist with analysis. As such, we employ a novel approach (Blok, et al, 2013) utilizing ChatGPT itself, to further inform our understanding of the situation. ChatGPT was selected as the genAI ’team member’ due to its ability to perform at deductive reasoning (Glickman, M., & Zhang, Y., 2024). Throughout this research, developing a team to include ChatGPT, influenced how the research team was structured. ChatGPT filled the role of ‘research assistant’, completing initial analysis after being given a prompt from the human, saving hours on initial findings which could then be verified and expanded on with literature and expertise. Once we assess the current impacts of genAI on knowledge work, we will then assess what this would look like from an organizational structure (based on current organizational structures and practices) with a view to understand the impacts to roles, followed by a snapshot review of what we know about genAI today to identify where future research could be focused with the topic of genAI impact on organizational structures. A view of likely near-term impacts – which are already momentous – will sensitize us to the possibility of even more extensive disruption in the medium and longer term. We provide a holistic view of the organizational changes that are likely to result from the uptake of genAI. An improved understanding of this will enable organizations to take steps to avoid the large-scale disruptions and organizational failures and prepare for this disruptive new technology. We also identify areas for future research to further our collective understanding on these impacts in a rapidly evolving environment. The following sections outline the research topic background, methodology, findings, discussion and conclusions. Following this, we reflect on completing data analysis utilizing ChatGPT to gather additional insights on the topic of AI impact on organizational structures. 2 Background and Related Work 2.1 GenAI and its impact on how organizations work Artificial intelligence is not a single technology, but a wide range of technologies that are developing independently, while also creating ‘combinatorial’ effects as independently developed capabilities interact to create new capabilities. There are many genAI technologies under development, in use in highly specialized contexts, for example, genAI is being investigated to boost the thinking skills of commanders of nuclear submarines, (Wilson, 2020), to read chest x-rays (King, 2018), and to support investment decisions (Ashta & Herrmann, 2021). Additionally, genAI is expanding the organizational toolkit with tools such as Copilot, being used, for example, in Excel for data analysis and developing artifacts in Word and PowerPoint (Spataro, 2023). The European Commission, in 2023, updated their taxonomy of genAI capability that includes reasoning (knowledge representation and automated reasoning); planning (planning and scheduling, searching, and optimization); machine learning; and natural language processing. We should assume that tools such as ChatGPT, while LLM-based, can draw on knowledge and extant capabilities in all these listed areas. There are many benefits to the introduction of genAI in organizations, such as the streamlining of activities like recruiting, through to loan assessments (Gil et al. 2019), integrating genAI causes competitive advantage (or defense against disruptive change) through enhancing strategic decision-making processes and increased performance, and reduces the requirement for menial tasks to be completed by humans through process automation, and can assist in customer relations and communication (Borges et al. 2021). However, there are also concerns relating to privacy of information, security of the systems, the current issue around genAI hallucinating outcomes (Birhane et al. 2023), bias of algorithms (Zhai, et al., 2024), data protection (Rezaei, et al., 2024) and the loss of knowledge associated to automating tasks (Soma, et al, 2022), which organizations need to be aware of when implementing and using genAI. GenAI is already being integrated into technologies (e.g., Microsoft’s Copilot, SAP’s Joule), regardless of the concerns and challenges currently existing, it will likely only become more integrated with our everyday lives. Some organizations are already embracing genAI to increase productivity, such as Akash Nigam, founder of Genies, who has purchased OpenAI subscriptions for all employees with the encouragement for employees to then use it in any, and every, aspect of their work (Mok, 2023). Strategically implementing such technologies is showing emerging themes of how roles, and work design, is changing through the adoption of genAI (Woodruff et al., 2024). 2.2 AI and organizational structures For most of the 20 th century, until the present day, Tayloristic “scientific” management based on division of labour has dominated thinking about organizational roles (Galpin, 2007). There are several variations on this theme, including hierarchical structures with multiple layers of management, and flat structures with few layers between management and front-line staff (shown in figure 1), as well as divisional structures based around locations, and network structures that emphasize horizontal relationships between colleagues at similar organizational levels, as well as relationships with supervising manager (Williams, 2023). In line with other approaches, lean management has contributed to focus more on process-oriented organizational structures instead of functions so that collaborations are formed according to their process assignments (Womack & Jones, 2003). Moreover, the role of independently acting employees and a collaboration is promoted for which information and communication technology enables a higher empowerment which means not to follow hierarchical orders only (Leyer, et al, 2019). Although there is agreement that genAI will be embedded, to some degree, in organizations in future, there is a lack of understanding around exactly which roles, the quantity of roles, and the severity of impact genAI could have on organizations when considered as a whole. Studies have been conducted on the likely impacts of genAI on specific functional areas, e.g., FinTech (Cao et al., 2021, Ashta & Herrman, 2021) and human resource management (Vrontis et al., 2022; Yawalkar, 2019). There is however limited evidence of more substantial research being completed on how genAI impacts organizational structures for core business functions, primarily through the implementation of technologies enhancing productivity and collaboration. Research is also seemingly in agreement that when organizations implement new technologies, they are disruptive, they create opportunities, risks, and challenges (Skog, et al, 2018). Skog et al. (2018) states, “When firms face the threat of digital disruption there is often an acute need to react due to the rapidity and systemic nature of environmental change along with diminishing business results”. We are already seeing the beginning of a genAI adoption wave with potential for widespread disruption. In order to analyze the technological impacts, we draw on Frey & Osborne’s (2017) research regarding the probabilities of which roles would be susceptible to computerization. Frey & Osborne stated, “Our model predicts that most workers in transportation and logistics occupations, together with the bulk of office and administrative support workers, and labour in production occupations, are at risk.” (Frey and Osborne 2017). Although Frey & Osborne (2017) do not directly address the impact these role changes will have on organizational structures, their research provides an opportunity for us to develop this understanding further and analyze the impacts in relation to genAI on organizational structures. Additionally, we can draw on important learnings such as Frey & Osborne’s closing statement “For workers to win the race, however, they will have to acquire creative and social skills” (Frey and Osborne, 2017), which is particularly relevant when assessing the change which genAI will bring to corporate organizations and how employees will need to adapt, and advance their human capabilities and competencies, in order to move with the impending changes rather than be replaced by genAI (Paschen, et al., 2020). Drawing these threads together, the ability of genAI is already in the public domain, ready to disrupt employees roles and organizational structures. Anecdotally, many individuals are already using ChatGPT, among other tools, to improve their productivity. However, the full extent of impact that can be expected, and the impact on organizational structures, has received little attention. 3 Method: Identifying GenAI Impact On Organizational Models Together With ChatGPT We take a hybrid approach, using ChatGPT and human analysis, to discover a holistic understanding of genAI’s impact on roles and structures. Chubb et al’s (2022) research into using AI within academic research processes identified a number of positive and negative outcomes. Outlining that caution should be expressed when using an AI tool within research due to learned biases and plagiarism. However, more positively, using AI in research allows for an interdisciplinary view which may otherwise not be achieved, additionally, ChatGPT has the ability to process data more efficiently than its human counterpart, reaching outcomes in an effective manner. By coupling AI analysis with human analysis, identification of patterns in data and predictions on future expectations are able to be made (Chubb et al., 2022). Clear rules and guidelines for the use of AI in research are being published by many journals (e.g., Business Horizons; Academy of Management; Elsevier) with a focus on; 1. what is appropriate use of AI in articles, and 2. how this should be disclosed to readers, as well as 3. whether ChatGPT and other AI tools can be considered an author. Informed by these guidelines, we utilize both the skills of humans and that of ChatGPT, acknowledging the restrictions of the counterparts, including biases, ChatGPT’s time bound training data (we used a version which had the last data update from June 2024 (OpenAI, 2025)), and the risk of hallucinated outputs (Birhane et al. 2023). To mitigate these risks and restrictions of ChatGPT, we appointed a human researcher as Team Lead and included a step in our methodology to assess human and ChatGPT analysis side-by-side to identify discrepancies in findings prior to drawing conclusions. We used a hybrid approach where both ChatGPT and human analysis was completed in parallel, including multiple analyses such as visual, data, and literature/grey-literature review to ensure a holistic review of the issue can be understood (Paré et al, 2015). This novel approach allowed a multi-faceted analysis of the data with the assistance of genAI to assist in our understanding of the problem, context, and implications for consideration against organizational structures, allowing for a unique perspective on the data. In order to ensure minimal risk of hallucinated results, a side-by-side comparison was completed between the ChatGPT outputs and our own human analysis (as described in the ‘Data Analysis’ phase in figure 2). ChatGPT was not used in any other elements of this research paper, other than the two steps relating to the query, listed alongside ‘ChatGPT Query’ in figure 2. Step one: Role Selection The O*Net, 2023 database was selected as it had a strong list of roles for use in our analysis and included skills associated to each role with the level required for delivery. The initial list contained 873 roles, we removed industries which were not our focus, including education, trade, hospitality and medical. To ensure all data would be of value, we removed any roles which had no skill data associated to it. This left us with 85 knowledge work roles for further assessment (see Appendix for full compiled table). Step two: ChatGPT Query Prior to any further human analysis, the 85 roles were provided to ChatGPT with the query “Based on what you know, can you please read the following roles and create a table, adding two filled in columns. Column one should be the role I provide; column two, which roles are likely to be replaced by generative AI; and column three, which skills specifically for the role will be replaced by generative AI”. Note: only the role column was given (see appendix). Step three: Organizational Structure Review In isolation to step two, an organizational structure review was completed by the human researchers. Four core structures were analysed using literature to determine common structures seen in corporate organizations today. The four structures were Divisional, Network, Hierarchical, and Flat. Hierarchical and Flat structures were more prevalent and selected for use in our analysis (shown earlier in figure 1). Step four (a, b, and c) was completed simultaneously to ensure findings, implications, recommendations and conclusions could be made. Step four (a): Visual Analysis All 85 roles were added to a full organizational structure (see appendix), bundled into business units to show managerial levels, then colour coded based on the output from ChatGPT on each role. Red reflected that genAI would fully replace the role, amber would partially replace the role, and green would not be affected by the implementation of genAI. This visual was then duplicated and all roles which would fully be replaced by genAI was removed from the structure (figure 4). Once step four (b) was completed, figure 4 was then expanded to draw out all skills which would be replaced by genAI and depicted as a team structure which outsources some skills to genAI directly. Assessment on managerial levels was completed between the two structures and assessed against step four (b) and (c). Step four (b): Data Analysis The O*NET, 2023 data was then analyzed on the basis of skill level required for skills which ChatGPT determined would be completed by genAI in the future. Roles which only had high skill levels (>4) required to complete their role were removed from the analysis as it has been assumed these would be more complex and would require some form of human intervention (this was the case for 70 of the 85 roles selected). The roles with the low-level skills were then tabled alongside the output from ChatGPT for those same roles to determine if the role would likely be replaced by genAI in the future and assess for correlation (as noted in table 1). 11 of the 15 roles aligned with ChatGPT and the skill level analysis, leaving 4 roles which showed discrepancy, discussed in the findings section. Step four (c): Literature/Article review As literature is in its infancy on the impacts of genAI on organizational structures, a multi-modal approach was taken, where both academic and grey literature was used. There were three core questions we sought to answer through this step to enhance our finding analysis: What does the literature say about AI impact on organizational structures? What is the media saying about concerns regarding impacts of AI on organizations? Do the reports and literature align with what the visual and data analysis depicts? Through performing targeted search on these questions, we were able to gain a more holistic understanding of the impacts we were seeing through our data and visual analysis. As step four (c) was completed in parallel with step four (a) and (b) we were able to be more targeted with our search on themes we saw emerging, for example, the move from hierarchical to flat, which has been a common move over recent times. This conceptual literature review is described by Paré et al, 2016, as one which assists in the understanding of attributes and basic elements of the concept. Our conceptual literature review of the listed questions, coupled with our own analysis, enhanced our hybrid approach of this research to further develop initial findings. Once step four was complete, implications, recommendations, and conclusions were made and documented. 4 Findings 4.1 GenAI and the change in roles While reviewing the skills ChatGPT identified as likely to be replaced by genAI, there is a common theme emerging. Many of the mundane and repetitive, collection and analysis type skills were identified as the most likely to be replaced (see appendix). There is also a theme emerging with technical ICT skills which are likely to be replaced, such as monitoring, triage, troubleshooting, and testing. When the skills anticipated to be replaced are compared to those listed in the source data (appended), these can be summarized below: Coordination Troubleshooting Operations analysis Monitoring Writing Programming Management of financial resources Management of material resources Quality control analysis Operations monitoring Technology design Assessing these skills against the roles identified, these impact all 85 roles, this is in contrast to ChatGPT’s list of 79 roles which will be partially or fully replaced, as all roles will be impacted by genAI’s ability to complete tasks with basic skills such as writing and editing tasks which was largely neglected from the information provided by ChatGPT. This discrepancy in our analysis and that of ChatGPT’s, could be as a result of the authors not providing enough context for ChatGPT to provide more thorough analysis, or as a result of bias or hallucination. Through this response, the authors acknowledge the importance of reviewing genAI outputs, applying critical thinking over results, along with expert knowledge where possible. Our data shows 70 roles which require high level of skills (rated >4). This information tells us that there is expected change for the 70 roles as the listed skills associated to them will be partially, if not entirely, replaced by genAI, and for the 15 roles which only required a low level of the skills (0-3.99), these are likely to no longer be performed by a human in the future. When we compare this to what ChatGPT tells us, it becomes clear how significant an impact this will have on organizations and how they are structured. However, we can also identify five roles where there is a discrepancy between the O*NET data and ChatGPT assessment. This shows that there is a possibility that Computer Network Support Specialists, Customer Service Representatives, First-Line Supervisors of Personal Service Workers, Production, Planning and Expediting Clerks, and Receptionist/Information Clerks will not entirely be replaced by genAI, even though the skills assessed say they will. The 15 roles which only required a low level of the skills assessed are listed below (Table 1) against the output from ChatGPT. Table 1 Low skill level roles comparison with ChatGPT response Role Low skill (<4) ChatGPT - will the role be replaced/augmented by AI Bill and Account Collectors Y YES Billing and Posting Clerks Y YES Bookkeeping, Accounting, and Auditing Clerks Y YES Computer Network Support Specialists Y PARTIALLY Customer Service Representatives Y PARTIALLY Data Entry Keyers Y YES Document Management Specialists Y YES File Clerks Y YES First-Line Supervisors of Personal Service Workers Y NO Office Clerks, General Y YES Payroll and Timekeeping Clerks Y YES Production, Planning, and Expediting Clerks Y PARTIALLY Receptionists and Information Clerks Y PARTIALLY Secretaries and Administrative Assistants, Except Legal, Medical, and Executive Y YES Word Processors and Typists Y YES 4.2 Changing roles leading to a different organizational structure Turning our focus now to the organizational structure impacts, Figure 3 shows what impact such changes could have on an organizational structure in the future, assuming ChatGPT’s predictions are accurate. Although every organization does not necessarily have every role we are assessing, for hypothetical clarity, the 85 roles have been included in a single snapshot to clearly articulate the scope of change. With 26 operational roles expected to be replaced by genAI, and a further 53 partially replace (middle management (13), operational (37), upper management (3)) we must turn our focus to the roles that will change due to the introduction of genAI, and mitigate the impact of those which will be removed. When removing the genAI replaced roles from the structure and adding in the skills which genAI will perform in future, based on our analysis, the scope of what we need to adapt to becomes visible. ChatGPT identified 79 skills which genAI will likely replace in the future; a sample of these are shown in the bottom right-hand box (“skills to be performed by AI”) in Figure 4. The changed roles (identified as yellow boxes) will need to adapt to a new structure where each role calls on genAI to perform 79 tasks, historically performed by a human. Although Figure 4 appears hierarchical, the structure is considered flat due to the reduction in the management levels as some skills are handed over to the genAI to complete. As we can see in the “skills to be performed by AI” box, this is a result of 79 named skills by ChatGPT that will be completed by genAI in future (sample of simplified skills shown in figure 4). These such skills, as identified earlier, remove 26 operational roles, and likely a reduction in management roles as a result of this. Structures will be “flatter” because they will not be as deep, with lower level knowledge-worker roles fully or partly replaced. 5 Discussion This research identifies implications for organizations due to the level of change required as a result of genAI being implemented into existing processes, roles and overarching organizational structures. These are not trivial changes that can easily be incorporated into existing structures and practices. The wave of high-profile business failures in the recent past, such as Kodak (Mui, 2012), and Xerox (Kulkarni, et al., 2020), were attributed not to a lack of knowledge about emerging digital technologies, but a failure to absorb these effectively into the organization (Mui, 2012). If we integrate the findings presented above with regard to organizational structures and job disruptions, with what we know about the current uptake of tools like ChatGPT, several things become clear. 5.1 “Coming ready or not” With more genAI tools appearing on the market, the effect on productivity is already being seen. Employees are actively sharing on platforms such as Reddit and TikTok how they are using genAI to streamline their work, allowing more time for themselves. Whether organizations want this or not, it is here now and it is impacting employee outputs. Individuals and organizations that are quick to redesign jobs and organization structures to incorporate genAI in a managed way, are likely to gain competitive advantage as productivity is enabled. Conversely organizations that do not have a coherent strategy for AI adoption into roles and if organizational reporting lines for genAI are not developed, they may experience uncontrolled use of AI by employees, with the potential for instability and political in-fighting as unplanned changes occur. The increase of performance resulting from the introduction of new digital technology to an organization is not a new concept. This is most commonly due to the introduction of new business processes and ‘ways of working’ which new technology implementation, and newly integrated technologies bring with it (Martinez-Caro, et al., 2020). Through innovation and cost reduction of processes, productivity is seemingly improved particularly in industries such as telecommunications and IT services, however without policies in place, this enhancement is mare perception (Pilat & Criscuolo, 2018). It is clear through research that without key capabilities and technology investment, productivity can be negatively impacted due to the lack of understanding and training of the new digital technologies (Pilat & Criscuolo, 2018) opening organizations to increased risk if proper adoption is not completed. 5.2 Unprepared organizations will be exposed to risk If individuals or groups within the organization incorporate genAI unilaterally into their jobs and structure, without an overriding plan, the organization will be exposed to a wide range of risks associated with genAI use. This may take many forms, for example mistakes made by genAI may not be spotted, genAI outputs may not be properly quality controlled or audited, and ethical, security, privacy, and copyright concerns may not be identified and managed (Pilat & Criscuolo, 2018). Without sufficient preparations and employee training, the implementation of genAI tools in an organization could be detrimental, exposing them to privacy and security risks, among many others. There are many people who are yet to understand that tools such as ChatGPT are not ‘thinking machines’, they are ‘predictive machines’ (Dodgson, 2023). The likes of genAI tools such as ChatGPT, which we utilized for analysis in this paper, are trained of information from a point in time (in our case, June 2024), without understanding this constraint, employees and organizations are likely to experience low quality outputs, and without understanding where the information has been drawn from, possible copywrite infringements. As the focus has shifted from new technology hype to the implementation of responsible AI, including a focus on human-centered technology, security and privacy of data (Vassilakopoulou, et al., 2022), organizational strategies and responsible AI policies must be put in place to protect communities while preparing for the use of genAI more widely throughout their organizations. In order to achieve this, as stated in Liu, et al.’s (2022) special issue editorial, the issue purpose was to “bring together researchers and practitioners working on network security and AI communities to present their recent researches and applications, and also to show how to seize opportunities and overcome challenges brought about by AI in security and privacy of emerging applications” (Liu, et al., 2022), however, there should be an ongoing collaboration between researchers and practitioners to ensure ongoing growth in creating more secure technologies for organizations, and society. 5.3 Changing roles As noted throughout our analysis, the implementation of genAI into roles will likely be the biggest impact on organizational structures . As this is a relatively new concept with minimal real-life examples to learn from, organizations are needing to strategize for many unknown impacts while navigating disruption to current employees and how they work today. Organizational culture plays a major part in the effectiveness of implementing new technologies (Martinez-Caro, et al., 2020). The introduction of genAIwill likely be no different. Having a strong organizational digital culture will enable successful implementation of such change, and in doing so propel the organization ahead of its competitors. Developing roles which incorporate both human and genAI elements is transformative, and adoption is moving quickly (Nyagadza, et al., 2022). Additionally, the way human’s problem-solve, and the way computers do, differ substantially, the strategy behind changing roles to include genAI capability needs to be calculated with this at the center of the strategy. Knowing exactly when and how to re-design roles and incorporating these into an organizational structure, and ensuring a strong digital culture is in place, is an opportunity to get ahead of competitors, but also a risk of major disruption to current business, or even failure. It is therefore imperative that organizations plan for a controlled integration of genAI tools through strategized adjustments to their organizational structure. This is particularly important when considering existing roles that will be significantly impacted by the change (as shown in figure 4). Organizational structuring will continue to flatten, but not necessarily be more process oriented. While prior ICT developments enabled a better information flow within organizations as a means to provide communication channels and access to information in predefined ways, genAI such as ChatGPT leads to an additional empowerment of individual employees. Every employee can access vast amount of data inside and outside the organization without fundamental training in the domain or deep experience in the organization. Formal languages as well as demands to IT departments to establish certain query options are not necessary. As such, ChatGPT enables a much more connected organization in which individuals and teams of individuals can work together with a decreased formal organizational structure. 6 Implications For Organizations To enable the successful implementation of significant technology change, organizational leaders need to ensure their own awareness and understanding of the changes required, clearly define, and strategize what will change, and how, and offer appropriate training and learning opportunities to employees to allow for genAI to be effectively adopted into roles. This includes knowing when to implement digital technology such as genAI (Holmstrom, 2022), minimizing impacts on employee wellbeing, while increasing productivity and profit opportunities. This may involve the following areas. 6.1 Proactive change management of organizational design . The large-scale changes that will occur or are presently occurring in job roles and organizational structures need to be actively managed and not left to evolve in an uncontrolled fashion. Organizations need to redesign specific job roles and their place in the organization structure proactively, as well as moving to a flatter organization structure with the implementation of genAI. 6.2 Upskilling through organizational redesign . Senior Managers will likely have received their professional training, education, and much of their professional experience before the advent of end-user genAI tools, and in a “conventional” more hierarchical organization. In order to enable success, they must be aware of the extent of change genAI tools will bring to their departments, and practice personal continuous learning to ensure they are able to support their human teams, and, where applicable, manage their genAI team members. Additionally, Senior Managers need the opportunity to build their own skills in order to redesign roles and organizational structures in their areas of expertise. It should also be the responsibility of organizational leaders to provide learning opportunities and training to all employees in their redesigned roles. This includes how to work in a collaborative environment with genAI, how to use the tool effectively, and developing clear responsibilities between what the human should retain and what the genAI can do (e.g., how to work together, as humans and AI, towards a common goal). 6.3 Identify new/changed roles (including new functions to partner with and supervise genAI) and incorporate them in the organization structure . It seemed inevitable, based on our analysis, that some new roles and functions may be required, as well as enhancements to existing roles. Our interactions with ChatGPT suggested that user experience, data analysis, solutions architecture, solution deployment, privacy and ethics, training, and performance evaluation roles would need to be enhanced to incorporate knowledge and skills in working with artificial intelligence (Jarrahi, 2018). 7 Conclusion Based on our results, we seriously need to consider what this means for our organizations at an individual level, and what we can do as leaders within organizations to ensure we embrace change and enable ongoing organizational success. From what we have seen through this research, there are 79 (out of 85) assessed corporate roles which will fully or partially be replaced by genAI, but the impact of this will stretch across the whole organization regardless of the final numbers. We need to prepare now for the significant changes genAI is likely to bring. Regarding when we will see the full effect of these changes within organizations, this is still up for debate, and depending on who you speak with, this can be anywhere from within the year, to the next decade. From ChatGPT’s perspective, this is also an unpredictable timeline - one has to keep in mind that the answers are based on the multitude of documents created by humans that are used in ChatGPT. Our paper has several theoretical implications: First, we provide a holistic view of the organizational changes that are likely to result from the uptake of genAI. Second, our results show how genAI can form collaborations with humans in different types of roles in organizations. The insights show that changes will be in line with ideas for organizational structuring from lean management. And third, we describe a new method including ChatGPT as a representative genAI on how to conduct analyses in organizational settings. This discussion with ChatGPT is only the start of what we need to know about how genAI will change organizations, and ultimately careers. There is opportunity for future research in many areas of this topic such as; what new roles will emerge as a result of genAI being embedded in organizations, the impacts to other industries such as medical and education, and how organizations can use this information to make strategic decisions around when and how to make significant changes to their organizations, such as introducing genAI in place of some human filled roles. There is still controversy around when, and how, fast genAI will change our organizations. However, as the predicted changes are already occurring and form part of the collective online body of knowledge about AI, and organizational change, organizations should not ignore these insights. To maintain their competitive position, recruit and retain employees who are proficient with new tools, and to upskill existing employees with extensive organizational knowledge, organizations need to understand what new roles and opportunities are arising, and pro-actively manage their job and organizational redesign. This can positively affect productivity (Wijayati et al. 2022) when implemented in an organization effectively. When asked about when we might start seeing these changes to roles, ChatGPT says we need to “stay informed about AI advancements and proactively prepare for the changing landscape through continuous learning and adaptability” (ChatGPT, 2023). 8 Reflection Of ChatGPT’s Role In This Paper This research was created in collaboration with Copilot, where results and analysis generated by ChatGPT was included with the analytical comparison, completed by the authors, against other data. We acknowledge that ChatGPT draws on a wide variety of text sources including news articles, Wikipedia and scientific journals, taken at a point in time. Therefore, we are aware there may be some differentiation in results over time and potential inaccuracies as a result of this method. Due to its ability to utilize LLM, ChatGPT provides a unique view on the question at hand, utilizing a wider range of literature in order to provide an answer in collaboration with human team members. We have used this data as a basis to commence the conversation on how such technologies can impact organizational structures, using the best efforts of existing AI technology, ChatGPT, to provide this understanding. As described by Yaroshenko & Iaroshenko (2023) some of the additional benefits experienced from utilizing ChatGPT as a valued team member for data analysis is the “versatility and time-saving features, ultimately leading to more impactful research outcomes” (pg. 197). Working with ChatGPT throughout the analysis of this paper was a unique experience, much like working with a human, we had to assess what ChatGPT’s strengths were and ensure effective communication was utilized for the best output. Some challenges we experienced included, the quantity of data ChatGPT could analyze at one time, meaning we had to run the same query for chunks of data inputs to ensure valuable results were given without ChatGPT leaving role analysis out (which was initially experienced prior to commencing the ChatGPT analysis portion). As discussed in our findings section, we also discovered discrepancies between ChatGPT’s analysis and our human analysis, in order to mitigate this, it was important that a human was appointed to oversee the results and critique the outputs. However, this unique experience also brought with it valuable outcomes. We believe by utilizing ChatGPT within the analysis portion of this research we have been enabled to provide a unique outlook on what changes we can experience within organizational structures, individual roles, and how genAI technologies may be able to be utilized in future research. ChatGPT was not part of the authorship of this paper. Declarations Author Contribution All authors contributed to the study conception and design. Material preparation, data collection and initial analysis was performed by CL. Analysis and findings were reviewed by AR, MT, ML. The first draft of the manuscript was written by CL and all authors revised and commented on previous versions of the manuscript. All authors read and approved the final manuscript. Data Availability Data is provided within the manuscript, as a link to Open Science Framework within the appendix section. An anonymous link has been created for blind review and will be public on acceptance.https://osf.io/rdm8p/?view_only=4ea1a3f39ff64c4ba421a0aa6c82d4d9 References Ashta, A., & Herrmann, H. (2021). “Artificial intelligence and fintech: An overview of opportunities and risks for banking, investments, and microfinance”. Strategic Change , 30(3), 211-222. Birhane, Abeba, Atoosa Kasirzadeh, David Leslie, and Sandra Wachter (2023). “Science in the Age of Large Language Models.” Nature Reviews Physics 5(5):277–80. doi: 10.1038/s42254-023-00581-4. Blok, K., M. Huijbregts, Lex Roes, Berthe van Haaster, M. K. Patel, E. Hertwich, M. Hauschild, P. Sellke, P. Antunes, S. Hellweg, A. Ciroth, and Mirjam Harmelink (2013). “A Novel Methodology for the Sustainability Impact Assessment of New Technologies.” Retrieved June 9, 2023 (https://dspace.library.uu.nl/handle/1874/303229). Borges, Aline F. S., Fernando J. B. Laurindo, Mauro M. Spínola, Rodrigo F. Gonçalves, and Claudia A. Mattos (2021). “The Strategic Use of Artificial Intelligence in the Digital Era: Systematic Literature Review and Future Research Directions.” International Journal of Information Management 57:102225. doi: 10.1016/j.ijinfomgt.2020.102225. Cao, L., Yang, Q., & Yu, P. S. (2021). “Data science and AI in FinTech: An overview”. International Journal of Data Science and Analytics , 12, 81-99. Chubb, J., Cowling, P., & Reed, D. (2022). Speeding up to keep up: Exploring the use of AI in the research process. AI & SOCIETY , 37 (4), 1439–1457. https://doi.org/10.1007/s00146-021-01259-0 Digital Government, New Zealand Government. (2023, July 26). What is Generative AI? Interim Generative AI Guidance for the Public Service. https://www.digital.govt.nz/standards-and-guidance/technology-and-architecture/artificial-intelligence/interim-generative-ai-guidance-for-the-public-service/what-is-generative-ai/ Dodgson, N. (2023, August 09). “Artificial intelligence: ChatGPT and human gullibility”. Policy Quarterly, 19:3, 19-24. European Commission (2023). "EU-U.S. Terminology and Taxonomy for Artificial Intelligence." (https://digital-strategy.ec.europa.eu/en/library/eu-us-terminology-and-taxonomy-artificial-intelligence). Frey, Carl Benedikt, and Michael A. Osborne (2017). “The Future of Employment: How Susceptible Are Jobs to Computerisation?” Technological Forecasting and Social Change 114:254–80. doi: 10.1016/j.techfore.2016.08.019. Galpin, T., Hilpirt, R., & Evans, B. (2007). The connected enterprise: beyond division of labor. Journal of Business Strategy, 28 (2), 38-47. García-Peñalvo, F., & Vázquez-Ingelmo, A. (2023). What Do We Mean by GenAI? A Systematic Mapping of The Evolution, Trends, and Techniques Involved in Generative AI. International Journal of Interactive Multimedia and Artificial Intelligence, 8(4), 7. https://doi.org/10.9781/ijimai.2023.07.006 Ghatak, A. (2023). Generative AI models like ChatGPT can be used to streamline IT operations. Dataquest, Retrieved from https://go.openathens.net/redirector/wgtn.ac.nz?url=https://www.proquest.com/trade-journals/generative-ai-models-like-chatgpt-can-be-used/docview/2811682650/se-2 Gil, Dario, Stacy Hobson, Aleksandra Mojsilovic, Ruchir Puri, and John Smith (2019). “AI for Management: An Overview | SpringerLink.” Retrieved June 3, 2023 (https://link.springer.com/chapter/10.1007/978-3-030-20680-2_1). Glickman, M.E., & Zhang, Y. (2024). "AI and Generative AI for Research Discovery and Summarization" Cornell University Library. e-ISSN: 2331-8422Hinsliff, G. (2023, May 4). “If Bosses Fail to Check AI's Onward March, Their Own Jobs Will Soon Be Written Out of The Script”. Retrieved from The Guardian: https://www.theguardian.com/commentisfree/2023/may/04/ai-jobs-script-machines-work-fun Holmstrom, J. (2022). “From AI to digital transformation: The AI readiness framework”. Business Horizons, 65 (3), 329-339. Jarrahi, M. H. (2018). "Artificial intelligence and the future of work: Human-AI symbiosis in organisational decision making". Business Horizons , 61(4), 577-586. Kaplan, A., & Haenlein, M. (2020). “Rulers of the world, unite! The challenges and opportunities of artificial intelligence”. Business Horizons , 63(1), 37-50. King, B. F. (2018). “Artificial intelligence and radiology: what will the future hold?” Journal of the American College of Radiology , 15(3), 501-503. Kulkarni, A., Dhongdi, A., Jadhav, S., Solankhe, P., Kashyap, A., & Hasbe, A. (2020, March 4). “Case Study on Xerox: Rise and fall of Xerox”. Retrieved from SlideShare: https://www.slideshare.net/SaurabhJadhav33/case-study-on-xerox-rise-and-fall-of-xerox Leyer, M., Richter, A., & Steinhuser, M. (2019). “Empowering shop floor workers with ICT. The role of participative designs”. International Journal of Operations and Production Management, 39 (1), 24-42. Liu, Q., Wang, G., Hu, J., Wu, J. (2022, March 17). “Preface of special issue on Artificial Intelligence: The security & privacy opportunities and challenges for emerging applications”. Future Generation Computer Systems, 133, 169-170. Marr, B. (2023, May 19). A Short History of ChatGPT: How We Got To Where We Are Today. Retrieved from Forbes: https://www.forbes.com/sites/bernardmarr/2023/05/19/a-short-history-of-chatgpt-how-we-got-to-where-we-are-today/?sh=2e97e227674f Marr, B. (2023, July 24). The Difference Between Generative AI And Traditional AI: An Easy Explanation For Anyone. Forbes. https://www.forbes.com/sites/bernardmarr/2023/07/24/the-difference-between-generative-ai-and-traditional-ai-an-easy-explanation-for-anyone/ Martinez-Caro, E., Cegarra-Navarro, J.G., Alfonso-Ruiz, F.J. (2020, February 15). “Digital technologies and firm performance: The role of digital organisational culture”. Technology Forecasting & Social Change, 154, 1-10. Mok, A. (2023, May 1). “A CEO is spending more than $2,000 a month on ChatGPT Plus accounts for all of his employees, and he says it's saving 'hours' of time”. Retrieved from Business Insider: https://www.businessinsider.com/ceo-buys-chatgpt-plus-accounts-all-employees-sees-productivity-boost-2023-5 Morris, M. R., Sohl-Dickstein, J., Fiedel, N., Warkentin, T., Dafoe, A., Faust, A., Farabet, C., & Legg, S. (2024). Position: Levels of AGI for Operationalizing Progress on the Path to AGI. ICML 2024 Conference. Mui, C. (2012, January 18). “How Kodak Failed”. Retrieved from Forbes: https://www.forbes.com/sites/chunkamui/2012/01/18/how-kodak-failed/?sh=5f3d5c156f27 Nyagadza, B., Pashapa, R., Chare, A., Mazuruse, G., Hove, P. K. (2022, January 05). “Digital technologies, Fourth Industrial Revolution (4IR) & Global Value Chains (GVCs) nexus with emerging economies’ future industrial innovation dynamics”. Cogent Economics & Finance, 10:1 , 1-14. O*NET (2023, May 9). “Occupational Data”. Retrieved from O*Net Resource Center: https://www.onetcenter.org/dictionary/27.2/excel/occupation_data.html OpenAI (2023). ChatGPT . Retrieved from Chat OpenAI: https://chat.openai.com Paré, G., Tate, M., Johnstone, D., & Kitsiou, S. (2016). Contextualizing the twin concepts of systematicity and transparency in information systems literature reviews. European Journal of Information Systems , 25 , 493-508. Paré, G., Trudel, M. C., Jaana, M., & Kitsiou, S. (2015). Synthesizing information systems knowledge: A typology of literature reviews. Information & Management , 52 (2), 183-199. Paschen, U., Pitt, C., & Kietzmann, J. (2020). “Artificial intelligence: Building blocks and an innovation typology”. Business Horizons, 63 (2), 147-155. Pilat, D., Criscuolo, C. (2018, August 12). “The future of productivity: What contribution can digital transformation make?”. Policy Quarterly, 14:3, 10-16. Press Trust of India (2023, June 5). "Around 4,000 Individuals Lost Their Jobs To Artificial Intelligence In May 2023: Report". Retrieved from Outlook: https://www.outlookindia.com/business/around-4000-individuals-lost-their-jobs-to-artificial-intelligence-in-may-2023-report-news-292234 Rahman, M., Terano, H. J. R., Rahman, N., Salamzadeh, A., & Rahaman, S. (2023). ChatGPT and Academic Research: A Review and Recommendations Based on Practical Examples. Journal of Education, Management and Development Studies , 3 (1), 1–12. https://doi.org/10.52631/jemds.v3i1.175 Rezaei, M., Pironti, M., & Quaglia, R. (2024). AI in knowledge sharing, which ethical challenges are raised in decision-making processes for organisations? Management Decision. https://doi.org/10.1108/MD-10-2023-2023 Schlegel, D., & Uenal, Y. (2021). A Perceived Risk Perspective on Narrow Artificial Intelligence. PACIS 2021 Proceedings, 44. https://aisel.aisnet.org/pacis2021/44 Skog, Daniel A., Henrik Wimelius, and Johan Sandberg (2018). “Digital Disruption.” Business & Information Systems Engineering, 60(5):431–37. doi: 10.1007/s12599-018-0550-4. Soma, Rebekka; Bratteteig, Tone; Saplacan, Diana; Schimmer, Robyn; Campano, Erik; and Verne, Guri B. (2022) "Strengthening Human Autonomy. In the era of autonomous technology," Scandinavian Journal of Information Systems , 34(2), Article 5. Available at: https://aisel.aisnet.org/sjis/vol34/iss2/5. Spataro, J. (2023, March 16). Introducing Microsoft 365 Copilot – your copilot for work. Official Microsoft Blog. https://blogs.microsoft.com/blog/2023/03/16/introducing-microsoft-365-copilot-your-copilot-for-work/ Vassilakopoulou, Polyxeni; Parmiggiani, Elena; Shollo, Arisa; and Grisot, Miria (2022) "Responsible AI:Concepts, critical perspectives and an Information Systems research agenda," Scandinavian Journal of Information Systems , 34(2), Article 3. Available at: https://aisel.aisnet.org/sjis/vol34/iss2/3 Vrontis, D., Christofi, M., Pereira, V., Tarba, S., Makrides, A., & Trichina, E. (2022). “Artificial intelligence, robotics, advanced technologies and human resource management: a systematic review”. The International Journal of Human Resource Management , 33(6), 1237-1266. Wijayati, Dewie Tri, Zainur Rahman, A’rasy Fahrullah, Muhammad Fajar Wahyudi Rahman, Ika Diyah Candra Arifah, and Achmad Kautsar (2022). “A Study of Artificial Intelligence on Employee Performance and Work Engagement: The Moderating Role of Change Leadership.” International Journal of Manpower 43(2):486–512. doi: 10.1108/IJM-07-2021-0423. Williams, S. (2023). “7 types of organizational structures (+ org charts for implementation)”. Retrieved from Lucidchart: https://www.lucidchart.com/blog/types-of-organizational-structures Wilson, C. (2020). “Artificial intelligence and warfare”. In M. Martellini, & R. Trapp, 21st Century Prometheus: Managing CBRN Safety and Security Affected by Cutting-Edge Technologies , 125-140. Womack, J.P. and Jones, D.T. (2003), “Lean Thinking. Banish Waste and Create Wealth in Your Corporation”, Free Press, New York, NY. Woodruff, A., Shelby, R., Kelley, P. G., Rousso-Schindler, S., Smith-Loud, J., & Wilcox, L. (2024). How Knowledge Workers Think Generative AI Will (Not) Transform Their Industries. Proceedings of the CHI Conference on Human Factors in Computing Systems, 1–26. https://doi.org/10.1145/3613904.3642700 Yaroshenko, T. O., & Iaroshenko, O. I. (2023). Artificial Intelligence (AI) for Research Lifecycle: Challenges and Opportunities. University Library at a New Stage of Social Communications Development. Conference Proceedings, 8, 194–201. https://doi.org/10.15802/unilib/2023_294639 Yawalkar, M. V. V. (2019). “A study of artificial intelligence and its role in human resource management”. International Journal of Research and Analytical Reviews (IJRAR) , 6(1), 20-24 Zhai, Y., Zhang, L., & Yu, M. (2024). AI in Human Resource Management: Literature Review and Research Implications. Journal of the Knowledge Economy. https://doi.org/10.1007/s13132-023-01631-z Additional Declarations No competing interests reported. Supplementary Files Appendix.docx Cite Share Download PDF Status: Posted Version 1 posted 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-6649663","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":463387808,"identity":"a17cb211-27a3-49dd-be35-1ccdbe4d2ef9","order_by":0,"name":"Chloe Latto","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4UlEQVRIiWNgGAWjYDACCeYGIGkBYjI+IFILI0iLBAMPGwOzAcla2CSI0sE/u7HxwQcGCXl7+eZj1bxthxn42w+wPfyBz5I7B5sNZzBIGPawsaXdBmmROJPAboDPPgOJxDZpHqDzeth4zG7ObLvNwHAD6EJ8vgJr+cMgYQ/SUgjSIg/SkkBIC9B9iSAtDB+BWgxAWg7g88uNxGbDHgOJ5J5jackSH8795zE8k9gm2YBHC/+M5IMPflTY2LY3Hz74IaEsTU7u+OFjkvhCDOo8BJMHmArw2TEKRsEoGAWjgBgAANnCQrHPCAsuAAAAAElFTkSuQmCC","orcid":"","institution":"Victoria University of Wellington","correspondingAuthor":true,"prefix":"","firstName":"Chloe","middleName":"","lastName":"Latto","suffix":""},{"id":463387809,"identity":"c4a9aa7a-c07b-4a57-9055-7cc730b6c8f2","order_by":1,"name":"Alexander Richter","email":"","orcid":"","institution":"Victoria University of Wellington","correspondingAuthor":false,"prefix":"","firstName":"Alexander","middleName":"","lastName":"Richter","suffix":""},{"id":463387810,"identity":"650278a9-4bd3-4cbc-9cfa-15cfe1ed5ac4","order_by":2,"name":"Mary Tate","email":"","orcid":"","institution":"Victoria University of Wellington","correspondingAuthor":false,"prefix":"","firstName":"Mary","middleName":"","lastName":"Tate","suffix":""},{"id":463387811,"identity":"4f71861e-8f77-4141-9c21-b65f0c80c2e0","order_by":3,"name":"Michael Leyer","email":"","orcid":"","institution":"Philipps University of Marburg","correspondingAuthor":false,"prefix":"","firstName":"Michael","middleName":"","lastName":"Leyer","suffix":""}],"badges":[],"createdAt":"2025-05-12 21:08:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6649663/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6649663/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":83667623,"identity":"ab0ca8f1-2a93-47a2-9bc4-b2e6d16e3c24","added_by":"auto","created_at":"2025-05-30 12:18:40","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":36124,"visible":true,"origin":"","legend":"\u003cp\u003eCommon functional organizational structures overview\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6649663/v1/529795845efe588ef590b373.png"},{"id":83667625,"identity":"895feb16-9bfa-4214-96c4-c6a404f2c09d","added_by":"auto","created_at":"2025-05-30 12:18:40","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":178969,"visible":true,"origin":"","legend":"\u003cp\u003eResearch and Analysis Method\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6649663/v1/de12136d01100245c1be7ff0.png"},{"id":83668214,"identity":"153795a0-6eed-4039-b31b-cefe968dd79c","added_by":"auto","created_at":"2025-05-30 12:26:40","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":30881,"visible":true,"origin":"","legend":"\u003cp\u003eFull Organizational Structure impact, as stated by ChatGPT\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6649663/v1/127fb09b20d55b9848e0227c.png"},{"id":83667628,"identity":"9ec13e3f-c67c-4f91-bd60-e1070900d4b1","added_by":"auto","created_at":"2025-05-30 12:18:40","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":285273,"visible":true,"origin":"","legend":"\u003cp\u003eOrganizational Structure, future state (based on ChatGPT predictions)\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6649663/v1/84c4f4aad415cb74935dcfa0.png"},{"id":83668494,"identity":"e6eaf7d3-34b1-40b2-a1be-71f7f8d6f6c3","added_by":"auto","created_at":"2025-05-30 12:34:41","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1300304,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6649663/v1/2a62a184-9793-4bcb-b07c-7116e6d211dc.pdf"},{"id":83667624,"identity":"0cdecfcc-fdd1-4494-a151-ad4d8cbb5db2","added_by":"auto","created_at":"2025-05-30 12:18:40","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":13922,"visible":true,"origin":"","legend":"","description":"","filename":"Appendix.docx","url":"https://assets-eu.researchsquare.com/files/rs-6649663/v1/42c099bf85d4d92a90aee853.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Generative AI’s Impact on Organizational Structures: An Analysis in Collaboration with ChatGPT","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eGenerative artificial intelligence (GenAI), is rapidly transforming organizational landscapes. Unlike earlier waves of AI and digitalization, which primarily targeted routine tasks, genAI is now poised to revolutionize areas previously untouched by automation, including those requiring basic creativity and nuanced communication (Ghatak, 2023). These advancements are not only reshaping individual roles but are also expected to drive significant changes in organizational structures. \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOrganizational structures ensure that responsibilities and functional areas are efficiently assigned to employees since an individual at the top of an organization (typically the CEO) cannot oversee everything in person. Thus, it is shared. Moreover, humans can only actively manage a certain number of social contacts, and personally interact with a limited number, so organizational units are required to ensure functionality of an organization. The introduction of genAI has enabled two main aspects in organizations. First, knowledge in the organization can be made easily accessible to every member of an organization, even if one has just started the job. Second, repetitive tasks can be completed in part by machines. As a result, more activities can be completed by genAI, with activities requiring human judgement remaining for humans (Hinsliff, 2023). However, to date, there has been no holistic analysis of the impacts of genAI on organization structures for core organizational functions, (Latto, et al., 2025), based on an analysis of its impacts on specific roles. Hence, in our study, we inquire how knowledge worker roles will change, re-designing work and how these changes will affect organizational structures.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eUtilizing AI tools is not just a current issue for knowledge work organizations; it is also becoming increasingly relevant in academic research. Tools such as ChatGPT, NotebookLM, and Microsoft\u0026rsquo;s Copilot provide new opportunities, but their emergence also raises important questions about the research process and ethics. While studies, including those summarized by Rahman et al. (2023), indicate that genAI is not yet suitable for conducting full empirical studies, they argue it can be effectively utilized for specific elements of the research process. In this paper, we explore this idea by integrating ChatGPT as a member of the research team alongside four human researchers to assist with analysis. As such, we employ a novel approach (Blok, et al, 2013) utilizing ChatGPT itself, to further inform our understanding of the situation. ChatGPT was selected as the genAI \u0026rsquo;team member\u0026rsquo; \u0026nbsp;due to its ability to perform at deductive reasoning (Glickman, M., \u0026amp; Zhang, Y., 2024). Throughout this research, developing a team to include ChatGPT, influenced how the research team was structured. ChatGPT filled the role of \u0026lsquo;research assistant\u0026rsquo;, completing initial analysis after being given a prompt from the human, saving hours on initial findings which could then be verified and expanded on with literature and expertise.\u003c/p\u003e\n\u003cp\u003eOnce we assess the current impacts of genAI on knowledge work, we will then assess what this would look like from an organizational structure (based on current organizational structures and practices) with a view to understand the impacts to roles, followed by a snapshot review of what we know about genAI today to identify where future research could be focused with the topic of genAI impact on organizational structures. A view of likely near-term impacts \u0026ndash; which are already momentous \u0026ndash; will sensitize us to the possibility of even more extensive disruption in the medium and longer term.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe provide a holistic view of the organizational changes that are likely to result from the uptake of genAI. An improved understanding of this will enable organizations to take steps to avoid the large-scale disruptions and organizational failures and prepare for this disruptive new technology. We also identify areas for future research to further our collective understanding on these impacts in a rapidly evolving environment.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe following sections outline the research topic background, methodology, findings, discussion and conclusions. Following this, we reflect on completing data analysis utilizing ChatGPT to gather additional insights on the topic of AI impact on organizational structures.\u003c/p\u003e"},{"header":"2 Background and Related Work","content":"\u003cp\u003e\u003cstrong\u003e2.1 GenAI and its impact on how organizations work\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eArtificial intelligence is not a single technology, but a wide range of technologies that are developing independently, while also creating \u0026lsquo;combinatorial\u0026rsquo; effects as independently developed capabilities interact to create new capabilities. There are many genAI technologies under development, in use in highly specialized contexts, for example, genAI is being investigated to boost the thinking skills of commanders of nuclear submarines, (Wilson, 2020), to read chest x-rays (King, 2018), and to support investment decisions (Ashta \u0026amp; Herrmann, 2021). Additionally, genAI is expanding the organizational toolkit with tools such as Copilot, being used, for example, in Excel for data analysis and developing artifacts in Word and PowerPoint (Spataro, 2023). The European Commission, in 2023, updated their taxonomy of genAI capability that includes reasoning (knowledge representation and automated reasoning); planning (planning and scheduling, searching, and optimization); machine learning; and natural language processing. We should assume that tools such as ChatGPT, while LLM-based, can draw on knowledge and extant capabilities in all these listed areas. There are many benefits to the introduction of genAI in organizations, such as the streamlining of activities like recruiting, through to loan assessments (Gil et al. 2019), integrating genAI causes competitive advantage (or defense against disruptive change) through enhancing strategic decision-making processes and increased performance, and reduces the requirement for menial tasks to be completed by humans through process automation, and can assist in customer relations and communication (Borges et al. 2021). However, there are also concerns relating to privacy of information, security of the systems, the current issue around genAI hallucinating outcomes (Birhane et al. 2023), bias of algorithms (Zhai, et al., 2024), data protection (Rezaei, et al., 2024) and the loss of knowledge associated to automating tasks (Soma, et al, 2022), which organizations need to be aware of when implementing and using genAI.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eGenAI is already being integrated into technologies (e.g., Microsoft\u0026rsquo;s Copilot, SAP\u0026rsquo;s Joule), regardless of the concerns and challenges currently existing, it will likely only become more integrated with our everyday lives. Some organizations are already embracing genAI to increase productivity, such as Akash Nigam, founder of Genies, who has purchased OpenAI subscriptions for all employees with the encouragement for employees to then use it in any, and every, aspect of their work (Mok, 2023). Strategically implementing such technologies is showing emerging themes of how roles, and work design, is changing through the adoption of genAI (Woodruff et al., 2024).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.2 AI and organizational structures\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFor most of the 20\u003csup\u003eth\u003c/sup\u003e century, until the present day, Tayloristic \u0026ldquo;scientific\u0026rdquo; management based on division of labour has dominated thinking about organizational roles (Galpin, 2007). There are several variations on this theme, including hierarchical structures with multiple layers of management, and flat structures with few layers between management and front-line staff (shown in figure 1), as well as divisional structures based around locations, and network structures that emphasize horizontal relationships between colleagues at similar organizational levels, as well as relationships with supervising manager (Williams, 2023). \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn line with other approaches, lean management has contributed to focus more on process-oriented organizational structures instead of functions so that collaborations are formed according to their process assignments (Womack \u0026amp; Jones, 2003). Moreover, the role of independently acting employees and a collaboration is promoted for which information and communication technology enables a higher empowerment which means not to follow hierarchical orders only (Leyer, et al, 2019).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAlthough there is agreement that genAI will be embedded, to some degree, in organizations in future, there is a lack of understanding around exactly which roles, the quantity of roles, and the severity of impact genAI could have on organizations when considered as a whole. Studies have been conducted on the likely impacts of genAI on specific functional areas, e.g., FinTech (Cao et al., 2021, Ashta \u0026amp; Herrman, 2021) and human resource management (Vrontis et al., 2022; Yawalkar, 2019). There is however limited evidence of more substantial research being completed on how genAI impacts organizational structures for core business functions, primarily through the implementation of technologies enhancing productivity and collaboration. Research is also seemingly in agreement that when organizations implement new technologies, they are disruptive, they create opportunities, risks, and challenges (Skog, et al, 2018). Skog et al. (2018) states, \u0026ldquo;When firms face the threat of digital disruption there is often an acute need to react due to the rapidity and systemic nature of environmental change along with diminishing business results\u0026rdquo;. We are already seeing the beginning of a genAI adoption wave with potential for widespread disruption.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn order to analyze the technological impacts, we draw on Frey \u0026amp; Osborne\u0026rsquo;s (2017) research regarding the probabilities of which roles would be susceptible to computerization. Frey \u0026amp; Osborne stated, \u0026ldquo;Our model predicts that most workers in transportation and logistics occupations, together with the bulk of office and administrative support workers, and labour in production occupations, are at risk.\u0026rdquo; (Frey and Osborne 2017). Although Frey \u0026amp; Osborne (2017) do not directly address the impact these role changes will have on organizational structures, their research provides an opportunity for us to develop this understanding further and analyze the impacts in relation to genAI on organizational structures. Additionally, we can draw on important learnings such as Frey \u0026amp; Osborne\u0026rsquo;s closing statement \u0026ldquo;For workers to win the race, however, they will have to acquire creative and social skills\u0026rdquo; (Frey and Osborne, 2017), which is particularly relevant when assessing the change which genAI will bring to corporate organizations and how employees will need to adapt, and advance their human capabilities and competencies, in order to move with the impending changes rather than be replaced by genAI (Paschen, et al., 2020).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDrawing these threads together, the ability of genAI is already in the public domain, ready to disrupt employees roles and organizational structures. Anecdotally, many individuals are already using ChatGPT, among other tools, to improve their productivity. However, the full extent of impact that can be expected, and the impact on organizational structures, has received little attention.\u0026nbsp;\u003c/p\u003e"},{"header":"3 Method: Identifying GenAI Impact On Organizational Models Together With ChatGPT","content":"\u003cp\u003eWe take a hybrid approach, using ChatGPT and human analysis, to discover a holistic understanding of genAI\u0026rsquo;s impact on roles and structures. Chubb et al\u0026rsquo;s (2022) research into using AI within academic research processes identified a number of positive and negative outcomes. Outlining that caution should be expressed when using an AI tool within research due to learned biases and plagiarism. However, more positively, using AI in research allows for an interdisciplinary view which may otherwise not be achieved, additionally, ChatGPT has the ability to process data more efficiently than its human counterpart, reaching outcomes in an effective manner. By coupling AI analysis with human analysis, identification of patterns in data and predictions on future expectations are able to be made (Chubb et al., 2022).\u003c/p\u003e\n\u003cp\u003eClear rules and guidelines for the use of AI in research are being published by many journals (e.g., Business Horizons; Academy of Management; Elsevier) with a focus on; 1. what is appropriate use of AI in articles, and 2. how this should be disclosed to readers, as well as 3. whether ChatGPT and other AI tools can be considered an author. Informed by these guidelines, we utilize both the skills of humans and that of ChatGPT, acknowledging the restrictions of the counterparts, including biases, ChatGPT\u0026rsquo;s time bound training data (we used a version which had the last data update from June 2024 (OpenAI, 2025)), and the risk of hallucinated outputs (Birhane et al. 2023). To mitigate these risks and restrictions of ChatGPT, we appointed a human researcher as Team Lead and included a step in our methodology to assess human and ChatGPT analysis side-by-side to identify discrepancies in findings prior to drawing conclusions.\u003c/p\u003e\n\u003cp\u003eWe used a hybrid approach where both ChatGPT and human analysis was completed in parallel, including multiple analyses such as visual, data, and literature/grey-literature review to ensure a holistic review of the issue can be understood (Par\u0026eacute; et al, 2015). This novel approach allowed a multi-faceted analysis of the data with the assistance of genAI to assist in our understanding of the problem, context, and implications for consideration against organizational structures, allowing for a unique perspective on the data.\u003c/p\u003e\n\u003cp\u003eIn order to ensure minimal risk of hallucinated results, a side-by-side comparison was completed between the ChatGPT outputs and our own human analysis (as described in the \u0026lsquo;Data Analysis\u0026rsquo; phase in figure 2). ChatGPT was not used in any other elements of this research paper, other than the two steps relating to the query, listed alongside \u0026lsquo;ChatGPT Query\u0026rsquo; in figure 2.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStep one:\u0026nbsp;\u003c/strong\u003eRole Selection\u003c/p\u003e\n\u003cp\u003eThe O*Net, 2023 database was selected as it had a strong list of roles for use in our analysis and included skills associated to each role with the level required for delivery. The initial list contained 873 roles, we removed industries which were not our focus, including education, trade, hospitality and medical. To ensure all data would be of value, we removed any roles which had no skill data associated to it. This left us with 85 knowledge work roles for further assessment (see Appendix for full compiled table).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStep two:\u0026nbsp;\u003c/strong\u003eChatGPT Query\u003c/p\u003e\n\u003cp\u003ePrior to any further human analysis, the 85 roles were provided to ChatGPT with the query \u0026ldquo;Based on what you know, can you please read the following roles and create a table, adding two filled in columns. Column one should be the role I provide; column two, which roles are likely to be replaced by generative AI; and column three, which skills specifically for the role will be replaced by generative AI\u0026rdquo;. Note: only the role column was given (see appendix).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStep three:\u0026nbsp;\u003c/strong\u003eOrganizational Structure Review\u003c/p\u003e\n\u003cp\u003eIn isolation to step two, an organizational structure review was completed by the human researchers. Four core structures were analysed using literature to determine common structures seen in corporate organizations today. The four structures were Divisional, Network, Hierarchical, and Flat. Hierarchical and Flat structures were more prevalent and selected for use in our analysis (shown earlier in figure 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eStep four (a, b, and c) was completed simultaneously to ensure findings, implications, recommendations and conclusions could be made.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStep four (a):\u0026nbsp;\u003c/strong\u003eVisual Analysis\u003c/p\u003e\n\u003cp\u003eAll 85 roles were added to a full organizational structure (see appendix), bundled into business units to show managerial levels, then colour coded based on the output from ChatGPT on each role. Red reflected that genAI would fully replace the role, amber would partially replace the role, and green would not be affected by the implementation of genAI. This visual was then duplicated and all roles which would fully be replaced by genAI was removed from the structure (figure 4). Once step four (b) was completed, figure 4 was then expanded to draw out all skills which would be replaced by genAI and depicted as a team structure which outsources some skills to genAI directly. Assessment on managerial levels was completed between the two structures and assessed against step four (b) and (c).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStep four (b):\u0026nbsp;\u003c/strong\u003eData Analysis\u003c/p\u003e\n\u003cp\u003eThe O*NET, 2023 data was then analyzed on the basis of skill level required for skills which ChatGPT determined would be completed by genAI in the future. Roles which only had high skill levels (\u0026gt;4) required to complete their role were removed from the analysis as it has been assumed these would be more complex and would require some form of human intervention (this was the case for 70 of the 85 roles selected). The roles with the low-level skills were then tabled alongside the output from ChatGPT for those same roles to determine if the role would likely be replaced by genAI in the future and assess for correlation (as noted in table 1). 11 of the 15 roles aligned with ChatGPT and the skill level analysis, leaving 4 roles which showed discrepancy, discussed in the findings section.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStep four (c):\u0026nbsp;\u003c/strong\u003eLiterature/Article review\u003c/p\u003e\n\u003cp\u003eAs literature is in its infancy on the impacts of genAI on organizational structures, a multi-modal approach was taken, where both academic and grey literature was used. There were three core questions we sought to answer through this step to enhance our finding analysis:\u003c/p\u003e\n\u003col start=\"1\" type=\"1\"\u003e\n \u003cli\u003eWhat does the literature say about AI impact on organizational structures?\u003c/li\u003e\n \u003cli\u003eWhat is the media saying about concerns regarding impacts of AI on organizations?\u003c/li\u003e\n \u003cli\u003eDo the reports and literature align with what the visual and data analysis depicts?\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eThrough performing targeted search on these questions, we were able to gain a more holistic understanding of the impacts we were seeing through our data and visual analysis. As step four (c) was completed in parallel with step four (a) and (b) we were able to be more targeted with our search on themes we saw emerging, for example, the move from hierarchical to flat, which has been a common move over recent times. This conceptual literature review is described by Par\u0026eacute; et al, 2016, as one which assists in the understanding of attributes and basic elements of the concept. Our conceptual literature review of the listed questions, coupled with our own analysis, enhanced our hybrid approach of this research to further develop initial findings.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOnce step four was complete, implications, recommendations, and conclusions were made and documented.\u003c/p\u003e"},{"header":"4 Findings","content":"\u003cp\u003e\u003cstrong\u003e4.1 GenAI and the change in roles\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWhile reviewing the skills ChatGPT identified as likely to be replaced by genAI, there is a common theme emerging. Many of the mundane and repetitive, collection and analysis type skills were identified as the most likely to be replaced (see appendix). There is also a theme emerging with technical ICT skills which are likely to be replaced, such as monitoring, triage, troubleshooting, and testing. When the skills anticipated to be replaced are compared to those listed in the source data (appended), these can be summarized below: \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003eCoordination\u003c/li\u003e\n \u003cli\u003eTroubleshooting\u003c/li\u003e\n \u003cli\u003eOperations analysis\u003c/li\u003e\n \u003cli\u003eMonitoring\u003c/li\u003e\n \u003cli\u003eWriting\u003c/li\u003e\n \u003cli\u003eProgramming\u003c/li\u003e\n \u003cli\u003eManagement of financial resources\u003c/li\u003e\n \u003cli\u003eManagement of material resources\u003c/li\u003e\n \u003cli\u003eQuality control analysis\u003c/li\u003e\n \u003cli\u003eOperations monitoring\u003c/li\u003e\n \u003cli\u003eTechnology design\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eAssessing these skills against the roles identified, these impact all 85 roles, this is in contrast to ChatGPT\u0026rsquo;s list of 79 roles which will be partially or fully replaced, as all roles will be impacted by genAI\u0026rsquo;s ability to complete tasks with basic skills such as writing and editing tasks which was largely neglected from the information provided by ChatGPT. This discrepancy in our analysis and that of ChatGPT\u0026rsquo;s, could be as a result of the authors not providing enough context for ChatGPT to provide more thorough analysis, or as a result of bias or hallucination. Through this response, the authors acknowledge the importance of reviewing genAI outputs, applying critical thinking over results, along with expert knowledge where possible. Our data shows 70 roles which require high level of skills (rated \u0026gt;4). This information tells us that there is expected change for the 70 roles as the listed skills associated to them will be partially, if not entirely, replaced by genAI, and for the 15 roles which only required a low level of the skills (0-3.99), these are likely to no longer be performed by a human in the future. When we compare this to what ChatGPT tells us, it becomes clear how significant an impact this will have on organizations and how they are structured. However, we can also identify five roles where there is a discrepancy between the O*NET data and ChatGPT assessment. This shows that there is a possibility that Computer Network Support Specialists, Customer Service Representatives, First-Line Supervisors of Personal Service Workers, Production, Planning and Expediting Clerks, and Receptionist/Information Clerks will not entirely be replaced by genAI, even though the skills assessed say they will.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe 15 roles which only required a low level of the skills assessed are listed below (Table 1) against the output from ChatGPT.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1\u003c/strong\u003e Low skill level roles comparison with ChatGPT response\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"623\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 349px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRole\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLow skill (\u0026lt;4)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 198px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eChatGPT - will the role be replaced/augmented by AI\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eBill and Account Collectors\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eBilling and Posting Clerks\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eBookkeeping, Accounting, and Auditing Clerks\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eComputer Network Support Specialists\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003ePARTIALLY\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eCustomer Service Representatives\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003ePARTIALLY\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eData Entry Keyers\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eDocument Management Specialists\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eFile Clerks\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eFirst-Line Supervisors of Personal Service Workers\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003eNO\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eOffice Clerks, General\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003ePayroll and Timekeeping Clerks\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eProduction, Planning, and Expediting Clerks\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003ePARTIALLY\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eReceptionists and Information Clerks\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003ePARTIALLY\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eSecretaries and Administrative Assistants, Except Legal, Medical, and Executive\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 349px;\"\u003e\n \u003cp\u003eWord Processors and Typists\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 76px;\"\u003e\n \u003cp\u003eY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 198px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003e4.2 Changing roles leading to a different organizational structure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTurning our focus now to the organizational structure impacts, Figure 3 shows what impact such changes could have on an organizational structure in the future, assuming ChatGPT\u0026rsquo;s predictions are accurate. Although every organization does not necessarily have every role we are assessing, for hypothetical clarity, the 85 roles have been included in a single snapshot to clearly articulate the scope of change.\u003c/p\u003e\n\u003cp\u003eWith 26 operational roles expected to be replaced by genAI, and a further 53 partially replace (middle management (13), operational (37), upper management (3)) we must turn our focus to the roles that will change due to the introduction of genAI, and mitigate the impact of those which will be removed. When removing the genAI replaced roles from the structure and adding in the skills which genAI will perform in future, based on our analysis, the scope of what we need to adapt to becomes visible. ChatGPT identified 79 skills which genAI will likely replace in the future; a sample of these are shown in the bottom right-hand box (\u0026ldquo;skills to be performed by AI\u0026rdquo;) in Figure 4. The changed roles (identified as yellow boxes) will need to adapt to a new structure where each role calls on genAI to perform 79 tasks, historically performed by a human.\u003c/p\u003e\n\u003cp\u003eAlthough Figure 4 appears hierarchical, the structure is considered flat due to the reduction in the management levels as some skills are handed over to the genAI to complete. As we can see in the \u0026ldquo;skills to be performed by AI\u0026rdquo; box, this is a result of 79 named skills by ChatGPT that will be completed by genAI in future (sample of simplified skills shown in figure 4). These such skills, as identified earlier, remove 26 operational roles, and likely a reduction in management roles as a result of this. Structures will be \u0026ldquo;flatter\u0026rdquo; because they will not be as deep, with lower level knowledge-worker roles fully or partly replaced.\u0026nbsp;\u003c/p\u003e"},{"header":"5 Discussion","content":"\u003cp\u003eThis research identifies implications for organizations due to the level of change required as a result of genAI being implemented into existing processes, roles and overarching organizational structures. These are not trivial changes that can easily be incorporated into existing structures and practices. The wave of high-profile business failures in the recent past, such as Kodak (Mui, 2012), and Xerox (Kulkarni, et al., 2020), were attributed not to a lack of knowledge about emerging digital technologies, but a failure to absorb these effectively into the organization (Mui, 2012). \u0026nbsp;If we integrate the findings presented above with regard to organizational structures and job disruptions, with what we know about the current uptake of tools like ChatGPT, several things become clear.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e5.1 \u0026ldquo;Coming ready or not\u0026rdquo;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWith more genAI tools appearing on the market, the effect on productivity is already being seen. Employees are actively sharing on platforms such as Reddit and TikTok how they are using genAI to streamline their work, allowing more time for themselves. Whether organizations want this or not, it is here now and it is impacting employee outputs. Individuals and organizations that are quick to redesign jobs and organization structures to incorporate genAI in a managed way, are likely to gain competitive advantage as productivity is enabled. Conversely organizations that do not have a coherent strategy for AI adoption into roles and if organizational reporting lines for genAI are not developed, they may experience uncontrolled use of AI by employees, with the potential for instability and political in-fighting as unplanned changes occur.\u003c/p\u003e\n\u003cp\u003eThe increase of performance resulting from the introduction of new digital technology to an organization is not a new concept. This is most commonly due to the introduction of new business processes and \u0026lsquo;ways of working\u0026rsquo; which new technology implementation, and newly integrated technologies bring with it (Martinez-Caro, et al., 2020). Through innovation and cost reduction of processes, productivity is seemingly improved particularly in industries such as telecommunications and IT services, however without policies in place, this enhancement is mare perception (Pilat \u0026amp; Criscuolo, 2018). It is clear through research that without key capabilities and technology investment, productivity can be negatively impacted due to the lack of understanding and training of the new digital technologies (Pilat \u0026amp; Criscuolo, 2018) opening organizations to increased risk if proper adoption is not completed.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e5.2 Unprepared organizations will be exposed to risk\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIf individuals or groups within the organization incorporate genAI unilaterally into their jobs and structure, without an\u0026nbsp;overriding plan, the organization will be exposed to a wide range of risks associated with genAI use. This may take many forms, for example mistakes made by genAI may not be spotted, genAI outputs may not be properly quality controlled or audited, and ethical, security, privacy, and copyright concerns may not be identified and managed (Pilat \u0026amp; Criscuolo, 2018).\u003c/p\u003e\n\u003cp\u003eWithout sufficient preparations and employee training, the implementation of genAI tools in an organization could be detrimental, exposing them to privacy and security risks, among many others. There are many people who are yet to understand that tools such as ChatGPT are not \u0026lsquo;thinking machines\u0026rsquo;, they are \u0026lsquo;predictive machines\u0026rsquo; (Dodgson, 2023). The likes of genAI tools such as ChatGPT, which we utilized for analysis in this paper, are trained of information from a point in time (in our case, June 2024), without understanding this constraint, employees and organizations are likely to experience low quality outputs, and without understanding where the information has been drawn from, possible copywrite infringements. As the focus has shifted from new technology hype to the implementation of responsible AI, including a focus on human-centered technology, security and privacy of data (Vassilakopoulou, et al., 2022), organizational strategies and responsible AI policies must be put in place to protect communities while preparing for the use of genAI more widely throughout their organizations. In order to achieve this, as stated in Liu, et al.\u0026rsquo;s (2022) special issue editorial, the issue purpose was to \u0026ldquo;bring together researchers and practitioners working on network security and AI communities to present their recent researches and applications, and also to show how to seize opportunities and overcome challenges brought about by AI in security and privacy of emerging applications\u0026rdquo; (Liu, et al., 2022), however, there should be an ongoing collaboration between researchers and practitioners to ensure ongoing growth in creating more secure technologies for organizations, and society.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e5.3 Changing roles\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAs noted throughout our analysis, the implementation of genAI into roles will likely be the biggest impact on organizational structures . As this is a relatively new concept with minimal real-life examples to learn from, organizations are needing to strategize for many unknown impacts while navigating disruption to current employees and how they work today. Organizational culture plays a major part in the effectiveness of implementing new technologies (Martinez-Caro, et al., 2020). The introduction of genAIwill likely be no different. Having a strong organizational digital culture will enable successful implementation of such change, and in doing so propel the organization ahead of its competitors. Developing roles which incorporate both human and genAI elements is transformative, and adoption is moving quickly (Nyagadza, et al., 2022). Additionally, the way human\u0026rsquo;s problem-solve, and the way computers do, differ substantially, the strategy behind changing roles to include genAI capability needs to be calculated with this at the center of the strategy. Knowing exactly when and how to re-design roles and incorporating these into an organizational structure, and ensuring a strong digital culture is in place, is an opportunity to get ahead of competitors, but also a risk of major disruption to current business, or even failure.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIt is therefore imperative that organizations plan for a controlled integration of genAI tools through strategized adjustments to their organizational structure. This is particularly important when considering existing roles that will be significantly impacted by the change (as shown in figure 4). Organizational structuring will continue to flatten, but not necessarily be more process oriented. While prior ICT developments enabled a better information flow within organizations as a means to provide communication channels and access to information in predefined ways, genAI such as ChatGPT leads to an additional empowerment of individual employees. Every employee can access vast amount of data inside and outside the organization without fundamental training in the domain or deep experience in the organization. Formal languages as well as demands to IT departments to establish certain query options are not necessary. As such, ChatGPT enables a much more connected organization in which individuals and teams of individuals can work together with a decreased formal organizational structure.\u003c/p\u003e"},{"header":"6 Implications For Organizations","content":"\u003cp\u003eTo enable the successful implementation of significant technology change, organizational leaders need to ensure their own awareness and understanding of the changes required, clearly define, and strategize what will change, and how, and offer appropriate training and learning opportunities to employees to allow for genAI to be effectively adopted into roles. This includes knowing when to implement digital technology such as genAI (Holmstrom, 2022), minimizing impacts on employee wellbeing, while increasing productivity and profit opportunities. This may involve the following areas.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e6.1 Proactive\u0026nbsp;change management of organizational design\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003eThe large-scale changes that will occur or are presently occurring in job roles and organizational structures need to be actively managed and not left to evolve in an uncontrolled fashion. Organizations need to redesign specific job roles and their place in the organization structure\u0026nbsp;proactively, as well as moving to a flatter organization structure with the implementation of genAI.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e6.2 Upskilling through organizational redesign\u003c/strong\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSenior Managers will likely have received their professional training, education, and much of their professional experience before the advent of end-user genAI tools, and in a \u0026ldquo;conventional\u0026rdquo; more hierarchical\u0026nbsp;organization. In order to enable success, they must be aware of the extent of change genAI tools will bring to their departments, and practice personal continuous learning to ensure they are able to support their human teams, and, where applicable, manage their genAI team members. Additionally, Senior Managers need the opportunity to build their own skills in order to redesign roles and organizational structures in their areas of expertise. It should also be the responsibility of organizational leaders to provide learning opportunities and training to all employees in their redesigned roles. This includes how to work in a collaborative environment with genAI, how to use the tool effectively, and developing clear responsibilities between what the human should retain and what the genAI can do (e.g., how to work together, as humans and AI, towards a common goal).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e6.3 Identify new/changed roles (including new functions to partner with and supervise genAI) and incorporate them in the organization structure\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003eIt seemed inevitable, based on our analysis, that some new roles and functions may be required, as well as enhancements to existing roles. Our interactions with ChatGPT suggested that user experience, data analysis, solutions architecture, solution deployment, privacy and ethics, training, and performance evaluation roles would need to be enhanced to incorporate knowledge and skills in working with artificial intelligence (Jarrahi, 2018).\u0026nbsp;\u003c/p\u003e"},{"header":"7 Conclusion","content":"\u003cp\u003eBased on our results, we seriously need to consider what this means for our organizations at an individual level, and what we can do as leaders within organizations to ensure we embrace change and enable ongoing organizational success. From what we have seen through this research, there are 79 (out of 85) assessed corporate roles which will fully or partially be replaced by genAI, but the impact of this will stretch across the whole organization regardless of the final numbers. We need to prepare now for the significant changes genAI is likely to bring. Regarding when we will see the full effect of these changes within organizations, this is still up for debate, and depending on who you speak with, this can be anywhere from within the year, to the next decade. From ChatGPT\u0026rsquo;s perspective, this is also an unpredictable timeline - one has to keep in mind that the answers are based on the multitude of documents created by humans that are used in ChatGPT.\u003c/p\u003e \u003cp\u003eOur paper has several theoretical implications: First, we provide a holistic view of the organizational changes that are likely to result from the uptake of genAI. Second, our results show how genAI can form collaborations with humans in different types of roles in organizations. The insights show that changes will be in line with ideas for organizational structuring from lean management. And third, we describe a new method including ChatGPT as a representative genAI on how to conduct analyses in organizational settings.\u003c/p\u003e \u003cp\u003eThis discussion with ChatGPT is only the start of what we need to know about how genAI will change organizations, and ultimately careers. There is opportunity for future research in many areas of this topic such as; what new roles will emerge as a result of genAI being embedded in organizations, the impacts to other industries such as medical and education, and how organizations can use this information to make strategic decisions around when and how to make significant changes to their organizations, such as introducing genAI in place of some human filled roles.\u003c/p\u003e \u003cp\u003eThere is still controversy around when, and how, fast genAI will change our organizations. However, as the predicted changes are already occurring and form part of the collective online body of knowledge about AI, and organizational change, organizations should not ignore these insights. To maintain their competitive position, recruit and retain employees who are proficient with new tools, and to upskill existing employees with extensive organizational knowledge, organizations need to understand what new roles and opportunities are arising, and pro-actively manage their job and organizational redesign. This can positively affect productivity (Wijayati et al. 2022) when implemented in an organization effectively. When asked about when we might start seeing these changes to roles, ChatGPT says we need to \u0026ldquo;stay informed about AI advancements and proactively prepare for the changing landscape through continuous learning and adaptability\u0026rdquo; (ChatGPT, 2023).\u003c/p\u003e"},{"header":"8 Reflection Of ChatGPT’s Role In This Paper","content":"\u003cp\u003eThis research was created in collaboration with Copilot, where results and analysis generated by ChatGPT was included with the analytical comparison, completed by the authors, against other data. We acknowledge that ChatGPT draws on a wide variety of text sources including news articles, Wikipedia and scientific journals, taken at a point in time. Therefore, we are aware there may be some differentiation in results over time and potential inaccuracies as a result of this method. Due to its ability to utilize LLM, ChatGPT provides a unique view on the question at hand, utilizing a wider range of literature in order to provide an answer in collaboration with human team members. We have used this data as a basis to commence the conversation on how such technologies can impact organizational structures, using the best efforts of existing AI technology, ChatGPT, to provide this understanding. As described by Yaroshenko \u0026amp; Iaroshenko (2023) some of the additional benefits experienced from utilizing ChatGPT as a valued team member for data analysis is the \u0026ldquo;versatility and time-saving features, ultimately leading to more impactful research outcomes\u0026rdquo; (pg. 197).\u003c/p\u003e\n\u003cp\u003eWorking with ChatGPT throughout the analysis of this paper was a unique experience, much like working with a human, we had to assess what ChatGPT\u0026rsquo;s strengths were and ensure effective communication was utilized for the best output. Some challenges we experienced included, the quantity of data ChatGPT could analyze at one time, meaning we had to run the same query for chunks of data inputs to ensure valuable results were given without ChatGPT leaving role analysis out (which was initially experienced prior to commencing the ChatGPT analysis portion). As discussed in our findings section, we also discovered discrepancies between ChatGPT\u0026rsquo;s analysis and our human analysis, in order to mitigate this, it was important that a human was appointed to oversee the results and critique the outputs.\u003c/p\u003e\n\u003cp\u003eHowever, this unique experience also brought with it valuable outcomes. We believe by utilizing ChatGPT within the analysis portion of this research we have been enabled to provide a unique outlook on what changes we can experience within organizational structures, individual roles, and how genAI technologies may be able to be utilized in future research.\u003c/p\u003e\n\u003cp\u003eChatGPT was not part of the authorship of this paper.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eAll authors contributed to the study conception and design. Material preparation, data collection and initial analysis was performed by CL. Analysis and findings were reviewed by AR, MT, ML. The first draft of the manuscript was written by CL and all authors revised and commented on previous versions of the manuscript. All authors read and approved the final manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eData is provided within the manuscript, as a link to Open Science Framework within the appendix section. An anonymous link has been created for blind review and will be public on acceptance.https://osf.io/rdm8p/?view_only=4ea1a3f39ff64c4ba421a0aa6c82d4d9\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAshta, A., \u0026amp; Herrmann, H. (2021). \u0026ldquo;Artificial intelligence and fintech: An overview of opportunities and risks for banking, investments, and microfinance\u0026rdquo;. \u003cem\u003eStrategic Change\u003c/em\u003e, 30(3), 211-222.\u003c/li\u003e\n\u003cli\u003eBirhane, Abeba, Atoosa Kasirzadeh, David Leslie, and Sandra Wachter (2023). \u0026ldquo;Science in the Age of Large Language Models.\u0026rdquo; \u003cem\u003eNature Reviews Physics\u003c/em\u003e 5(5):277\u0026ndash;80. doi: 10.1038/s42254-023-00581-4.\u003c/li\u003e\n\u003cli\u003eBlok, K., M. Huijbregts, Lex Roes, Berthe van Haaster, M. K. Patel, E. Hertwich, M. Hauschild, P. Sellke, P. Antunes, S. Hellweg, A. Ciroth, and Mirjam Harmelink (2013). \u0026ldquo;A Novel Methodology for the Sustainability Impact Assessment of New Technologies.\u0026rdquo; Retrieved June 9, 2023 (https://dspace.library.uu.nl/handle/1874/303229).\u003c/li\u003e\n\u003cli\u003eBorges, Aline F. S., Fernando J. B. Laurindo, Mauro M. Sp\u0026iacute;nola, Rodrigo F. Gon\u0026ccedil;alves, and Claudia A. Mattos (2021). \u0026ldquo;The Strategic Use of Artificial Intelligence in the Digital Era: Systematic Literature Review and Future Research Directions.\u0026rdquo; \u003cem\u003eInternational Journal of Information Management\u003c/em\u003e 57:102225. doi: 10.1016/j.ijinfomgt.2020.102225.\u003c/li\u003e\n\u003cli\u003eCao, L., Yang, Q., \u0026amp; Yu, P. S. (2021). \u0026ldquo;Data science and AI in FinTech: An overview\u0026rdquo;. \u003cem\u003eInternational Journal of Data Science and Analytics\u003c/em\u003e, 12, 81-99.\u003c/li\u003e\n\u003cli\u003eChubb, J., Cowling, P., \u0026amp; Reed, D. (2022). Speeding up to keep up: Exploring the use of AI in the research process. \u003cem\u003eAI \u0026amp; SOCIETY\u003c/em\u003e, \u003cem\u003e37\u003c/em\u003e(4), 1439\u0026ndash;1457. https://doi.org/10.1007/s00146-021-01259-0\u003c/li\u003e\n\u003cli\u003eDigital Government, New Zealand Government. (2023, July 26). What is Generative AI? Interim Generative AI Guidance for the Public Service. https://www.digital.govt.nz/standards-and-guidance/technology-and-architecture/artificial-intelligence/interim-generative-ai-guidance-for-the-public-service/what-is-generative-ai/\u003c/li\u003e\n\u003cli\u003eDodgson, N. (2023, August 09). \u0026ldquo;Artificial intelligence: ChatGPT and human gullibility\u0026rdquo;. \u003cem\u003ePolicy Quarterly, 19:3,\u003c/em\u003e 19-24.\u003c/li\u003e\n\u003cli\u003eEuropean Commission (2023). \u0026quot;EU-U.S. Terminology and Taxonomy for Artificial Intelligence.\u0026quot; (https://digital-strategy.ec.europa.eu/en/library/eu-us-terminology-and-taxonomy-artificial-intelligence).\u003c/li\u003e\n\u003cli\u003eFrey, Carl Benedikt, and Michael A. Osborne (2017). \u0026ldquo;The Future of Employment: How Susceptible Are Jobs to Computerisation?\u0026rdquo; \u003cem\u003eTechnological Forecasting and Social Change\u003c/em\u003e 114:254\u0026ndash;80. doi: 10.1016/j.techfore.2016.08.019.\u003c/li\u003e\n\u003cli\u003eGalpin, T., Hilpirt, R., \u0026amp; Evans, B. (2007). The connected enterprise: beyond division of labor. \u003cem\u003eJournal of Business Strategy, 28\u003c/em\u003e(2), 38-47.\u003c/li\u003e\n\u003cli\u003eGarc\u0026iacute;a-Pe\u0026ntilde;alvo, F., \u0026amp; V\u0026aacute;zquez-Ingelmo, A. (2023). What Do We Mean by GenAI? A Systematic Mapping of The Evolution, Trends, and Techniques Involved in Generative AI. International Journal of Interactive Multimedia and Artificial Intelligence, 8(4), 7. https://doi.org/10.9781/ijimai.2023.07.006\u003c/li\u003e\n\u003cli\u003eGhatak, A. (2023). Generative AI models like ChatGPT can be used to streamline IT operations. Dataquest, Retrieved from https://go.openathens.net/redirector/wgtn.ac.nz?url=https://www.proquest.com/trade-journals/generative-ai-models-like-chatgpt-can-be-used/docview/2811682650/se-2\u003c/li\u003e\n\u003cli\u003eGil, Dario, Stacy Hobson, Aleksandra Mojsilovic, Ruchir Puri, and John Smith (2019). \u0026ldquo;AI for Management: An Overview | SpringerLink.\u0026rdquo; Retrieved June 3, 2023 (https://link.springer.com/chapter/10.1007/978-3-030-20680-2_1).\u003c/li\u003e\n\u003cli\u003eGlickman, M.E., \u0026amp; Zhang, Y. (2024). \u0026quot;AI and Generative AI for Research Discovery and Summarization\u0026quot; Cornell University Library. e-ISSN: 2331-8422Hinsliff, G. (2023, May 4). \u0026ldquo;If Bosses Fail to Check AI\u0026apos;s Onward March, Their Own Jobs Will Soon Be Written Out of The Script\u0026rdquo;. Retrieved from The Guardian: https://www.theguardian.com/commentisfree/2023/may/04/ai-jobs-script-machines-work-fun\u003c/li\u003e\n\u003cli\u003eHolmstrom, J. (2022). \u0026ldquo;From AI to digital transformation: The AI readiness framework\u0026rdquo;. \u003cem\u003eBusiness Horizons, 65\u003c/em\u003e(3), 329-339.\u003c/li\u003e\n\u003cli\u003eJarrahi, M. H. (2018). \u0026quot;Artificial intelligence and the future of work: Human-AI symbiosis in organisational decision making\u0026quot;. \u003cem\u003eBusiness Horizons\u003c/em\u003e, 61(4), 577-586.\u003c/li\u003e\n\u003cli\u003eKaplan, A., \u0026amp; Haenlein, M. (2020). \u0026ldquo;Rulers of the world, unite! The challenges and opportunities of artificial intelligence\u0026rdquo;. \u003cem\u003eBusiness Horizons\u003c/em\u003e, 63(1), 37-50.\u003c/li\u003e\n\u003cli\u003eKing, B. F. (2018). \u0026ldquo;Artificial intelligence and radiology: what will the future hold?\u0026rdquo; \u003cem\u003eJournal \u003c/em\u003e\u003cem\u003eof the American College of Radiology\u003c/em\u003e, 15(3), 501-503.\u003c/li\u003e\n\u003cli\u003eKulkarni, A., Dhongdi, A., Jadhav, S., Solankhe, P., Kashyap, A., \u0026amp; Hasbe, A. (2020, March 4). \u0026ldquo;Case Study on Xerox: Rise and fall of Xerox\u0026rdquo;. Retrieved from SlideShare: https://www.slideshare.net/SaurabhJadhav33/case-study-on-xerox-rise-and-fall-of-xerox\u003c/li\u003e\n\u003cli\u003eLeyer, M., Richter, A., \u0026amp; Steinhuser, M. (2019). \u0026ldquo;Empowering shop floor workers with ICT. The role of participative designs\u0026rdquo;. \u003cem\u003eInternational Journal of Operations and Production Management, 39\u003c/em\u003e(1), 24-42.\u003c/li\u003e\n\u003cli\u003eLiu, Q., Wang, G., Hu, J., Wu, J. (2022, March 17). \u0026ldquo;Preface of special issue on Artificial Intelligence: The security \u0026amp; privacy opportunities and challenges for emerging applications\u0026rdquo;. \u003cem\u003eFuture Generation Computer Systems, 133,\u003c/em\u003e 169-170.\u003c/li\u003e\n\u003cli\u003eMarr, B. (2023, May 19). A Short History of ChatGPT: How We Got To Where We Are Today. Retrieved from Forbes: https://www.forbes.com/sites/bernardmarr/2023/05/19/a-short-history-of-chatgpt-how-we-got-to-where-we-are-today/?sh=2e97e227674f\u003c/li\u003e\n\u003cli\u003eMarr, B. (2023, July 24). The Difference Between Generative AI And Traditional AI: An Easy Explanation For Anyone. Forbes. https://www.forbes.com/sites/bernardmarr/2023/07/24/the-difference-between-generative-ai-and-traditional-ai-an-easy-explanation-for-anyone/\u003c/li\u003e\n\u003cli\u003eMartinez-Caro, E., Cegarra-Navarro, J.G., Alfonso-Ruiz, F.J. (2020, February 15). \u0026ldquo;Digital technologies and firm performance: The role of digital organisational culture\u0026rdquo;. Technology Forecasting \u0026amp; Social Change, 154, 1-10.\u003c/li\u003e\n\u003cli\u003eMok, A. (2023, May 1). \u0026ldquo;A CEO is spending more than $2,000 a month on ChatGPT Plus accounts for all of his employees, and he says it\u0026apos;s saving \u0026apos;hours\u0026apos; of time\u0026rdquo;. Retrieved from Business Insider: https://www.businessinsider.com/ceo-buys-chatgpt-plus-accounts-all-employees-sees-productivity-boost-2023-5\u003c/li\u003e\n\u003cli\u003eMorris, M. R., Sohl-Dickstein, J., Fiedel, N., Warkentin, T., Dafoe, A., Faust, A., Farabet, C., \u0026amp; Legg, S. (2024). Position: Levels of AGI for Operationalizing Progress on the Path to AGI. ICML 2024 Conference.\u003c/li\u003e\n\u003cli\u003eMui, C. (2012, January 18). \u0026ldquo;How Kodak Failed\u0026rdquo;. Retrieved from Forbes: https://www.forbes.com/sites/chunkamui/2012/01/18/how-kodak-failed/?sh=5f3d5c156f27\u003c/li\u003e\n\u003cli\u003eNyagadza, B., Pashapa, R., Chare, A., Mazuruse, G., Hove, P. K. (2022, January 05). \u0026ldquo;Digital technologies, Fourth Industrial Revolution (4IR) \u0026amp; Global Value Chains (GVCs) nexus with emerging economies\u0026rsquo; future industrial innovation dynamics\u0026rdquo;. \u003cem\u003eCogent Economics \u0026amp; Finance, 10:1\u003c/em\u003e, 1-14.\u003c/li\u003e\n\u003cli\u003eO*NET (2023, May 9). \u0026ldquo;Occupational Data\u0026rdquo;. Retrieved from O*Net Resource Center: https://www.onetcenter.org/dictionary/27.2/excel/occupation_data.html\u003c/li\u003e\n\u003cli\u003eOpenAI (2023). \u003cem\u003eChatGPT\u003c/em\u003e. Retrieved from Chat OpenAI: https://chat.openai.com\u003c/li\u003e\n\u003cli\u003ePar\u0026eacute;, G., Tate, M., Johnstone, D., \u0026amp; Kitsiou, S. (2016). Contextualizing the twin concepts of systematicity and transparency in information systems literature reviews. \u003cem\u003eEuropean Journal of Information Systems\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e, 493-508.\u003c/li\u003e\n\u003cli\u003ePar\u0026eacute;, G., Trudel, M. C., Jaana, M., \u0026amp; Kitsiou, S. (2015). Synthesizing information systems knowledge: A typology of literature reviews. \u003cem\u003eInformation \u0026amp; Management\u003c/em\u003e, \u003cem\u003e52\u003c/em\u003e(2), 183-199.\u003c/li\u003e\n\u003cli\u003ePaschen, U., Pitt, C., \u0026amp; Kietzmann, J. (2020). \u0026ldquo;Artificial intelligence: Building blocks and an innovation typology\u0026rdquo;. \u003cem\u003eBusiness Horizons, 63\u003c/em\u003e(2), 147-155.\u003c/li\u003e\n\u003cli\u003ePilat, D., Criscuolo, C. (2018, August 12). \u0026ldquo;The future of productivity: What contribution can digital transformation make?\u0026rdquo;. \u003cem\u003ePolicy Quarterly, 14:3,\u003c/em\u003e 10-16.\u003c/li\u003e\n\u003cli\u003ePress Trust of India (2023, June 5). \u0026quot;Around 4,000 Individuals Lost Their Jobs To Artificial Intelligence In May 2023: Report\u0026quot;. Retrieved from Outlook: https://www.outlookindia.com/business/around-4000-individuals-lost-their-jobs-to-artificial-intelligence-in-may-2023-report-news-292234\u003c/li\u003e\n\u003cli\u003eRahman, M., Terano, H. J. R., Rahman, N., Salamzadeh, A., \u0026amp; Rahaman, S. (2023). ChatGPT and Academic Research: A Review and Recommendations Based on Practical Examples. \u003cem\u003eJournal of Education, Management and Development Studies\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e(1), 1\u0026ndash;12. https://doi.org/10.52631/jemds.v3i1.175\u003c/li\u003e\n\u003cli\u003eRezaei, M., Pironti, M., \u0026amp; Quaglia, R. (2024). AI in knowledge sharing, which ethical challenges are raised in decision-making processes for organisations? Management Decision. https://doi.org/10.1108/MD-10-2023-2023\u003c/li\u003e\n\u003cli\u003eSchlegel, D., \u0026amp; Uenal, Y. (2021). A Perceived Risk Perspective on Narrow Artificial Intelligence. PACIS 2021 Proceedings, 44. https://aisel.aisnet.org/pacis2021/44\u003c/li\u003e\n\u003cli\u003eSkog, Daniel A., Henrik Wimelius, and Johan Sandberg (2018). \u0026ldquo;Digital Disruption.\u0026rdquo; Business \u0026amp; Information Systems Engineering, 60(5):431\u0026ndash;37. doi: 10.1007/s12599-018-0550-4.\u003c/li\u003e\n\u003cli\u003eSoma, Rebekka; Bratteteig, Tone; Saplacan, Diana; Schimmer, Robyn; Campano, Erik; and Verne, Guri B. (2022) \u0026quot;Strengthening Human Autonomy. In the era of autonomous technology,\u0026quot; \u003cem\u003eScandinavian Journal of Information Systems\u003c/em\u003e, 34(2), Article 5. Available at: https://aisel.aisnet.org/sjis/vol34/iss2/5.\u003c/li\u003e\n\u003cli\u003eSpataro, J. (2023, March 16). Introducing Microsoft 365 Copilot \u0026ndash; your copilot for work. Official Microsoft Blog. https://blogs.microsoft.com/blog/2023/03/16/introducing-microsoft-365-copilot-your-copilot-for-work/\u003c/li\u003e\n\u003cli\u003eVassilakopoulou, Polyxeni; Parmiggiani, Elena; Shollo, Arisa; and Grisot, Miria (2022) \u0026quot;Responsible AI:Concepts, critical perspectives and an Information Systems research agenda,\u0026quot; \u003cem\u003eScandinavian Journal of Information Systems\u003c/em\u003e, 34(2), Article 3. Available at: https://aisel.aisnet.org/sjis/vol34/iss2/3 \u003c/li\u003e\n\u003cli\u003eVrontis, D., Christofi, M., Pereira, V., Tarba, S., Makrides, A., \u0026amp; Trichina, E. (2022). \u0026ldquo;Artificial intelligence, robotics, advanced technologies and human resource management: a systematic review\u0026rdquo;. \u003cem\u003eThe International Journal of Human Resource Management\u003c/em\u003e, 33(6), 1237-1266.\u003c/li\u003e\n\u003cli\u003eWijayati, Dewie Tri, Zainur Rahman, A\u0026rsquo;rasy Fahrullah, Muhammad Fajar Wahyudi Rahman, Ika Diyah Candra Arifah, and Achmad Kautsar (2022). \u0026ldquo;A Study of Artificial Intelligence on Employee Performance and Work Engagement: The Moderating Role of Change Leadership.\u0026rdquo; \u003cem\u003eInternational Journal of Manpower\u003c/em\u003e 43(2):486\u0026ndash;512. doi: 10.1108/IJM-07-2021-0423.\u003c/li\u003e\n\u003cli\u003eWilliams, S. (2023). \u0026ldquo;7 types of organizational structures (+ org charts for implementation)\u0026rdquo;. Retrieved from Lucidchart: https://www.lucidchart.com/blog/types-of-organizational-structures\u003c/li\u003e\n\u003cli\u003eWilson, C. (2020). \u0026ldquo;Artificial intelligence and warfare\u0026rdquo;. In M. Martellini, \u0026amp; R. Trapp, \u003cem\u003e21st Century Prometheus: Managing CBRN Safety and Security Affected by Cutting-Edge Technologies\u003c/em\u003e, 125-140.\u003c/li\u003e\n\u003cli\u003eWomack, J.P. and Jones, D.T. (2003), \u0026ldquo;Lean Thinking. Banish Waste and Create Wealth in Your Corporation\u0026rdquo;, Free Press, New York, NY.\u003c/li\u003e\n\u003cli\u003eWoodruff, A., Shelby, R., Kelley, P. G., Rousso-Schindler, S., Smith-Loud, J., \u0026amp; Wilcox, L. (2024). How Knowledge Workers Think Generative AI Will (Not) Transform Their Industries. Proceedings of the CHI Conference on Human Factors in Computing Systems, 1\u0026ndash;26. https://doi.org/10.1145/3613904.3642700\u003c/li\u003e\n\u003cli\u003eYaroshenko, T. O., \u0026amp; Iaroshenko, O. I. (2023). Artificial Intelligence (AI) for Research Lifecycle: Challenges and Opportunities. University Library at a New Stage of Social Communications Development. Conference Proceedings, 8, 194\u0026ndash;201. https://doi.org/10.15802/unilib/2023_294639\u003c/li\u003e\n\u003cli\u003eYawalkar, M. V. V. (2019). \u0026ldquo;A study of artificial intelligence and its role in human resource management\u0026rdquo;. \u003cem\u003eInternational Journal of Research and Analytical Reviews (IJRAR)\u003c/em\u003e, 6(1), 20-24\u003c/li\u003e\n\u003cli\u003eZhai, Y., Zhang, L., \u0026amp; Yu, M. (2024). AI in Human Resource Management: Literature Review and Research Implications. Journal of the Knowledge Economy. https://doi.org/10.1007/s13132-023-01631-z\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Generative Artificial Intelligence, Organizational Structures, ChatGPT, Re-designing Roles","lastPublishedDoi":"10.21203/rs.3.rs-6649663/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6649663/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eArtificial intelligence has been rapidly increasing in sophistication and impact, reinforcing discussions of its extensive potential to disrupt jobs and organizations. This paper investigates the disruptive potential of generative artificial intelligence (genAI) in organizations by investigating its impact on organizational structures for core business functions as work re-design occurs with the introduction of genAI. Our analysis investigates the impacts of genAI, but we also use genAI as a partner, in collaboration with the authors. Data from the O*Net 2023 database is utilized to consolidate a list of 85 corporate knowledge worker roles and (associated skill levels), as part of our analysis, ChatGPT contributed its assessment as to which roles genAI will take over partly or fully in the near future. Our analysis led to a \u0026ldquo;flat\u0026rdquo; organization structure which allowed the surfacing of the extent of impact genAI will have on roles and, in turn, organizational structures. Results suggest that use of genAI is accelerating the trend of flattening hierarchies and will lead to more independent teams with joint Human-AI capability. Finally, we observe that AI is already on a disruptive trajectory. Individuals and organizations not engaging with the potential of genAI, and the expected organizational changes, are at serious risk of underperformance, role redundancy, and negative outcomes from the uncontrolled use of genAI.\u003c/p\u003e","manuscriptTitle":"Generative AI’s Impact on Organizational Structures: An Analysis in Collaboration with ChatGPT","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-30 12:18:35","doi":"10.21203/rs.3.rs-6649663/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"65befe45-0474-4683-b29c-6e9fe4e9f0ad","owner":[],"postedDate":"May 30th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-05-19T10:09:26+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-30 12:18:35","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6649663","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6649663","identity":"rs-6649663","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}