Can AI Teach Sex Ed? A Systematic Review of the Use of Artificial Intelligence in Sexual and Reproductive Health Education

Can AI Teach Sex Ed? A Systematic Review of the Use of Artificial Intelligence in Sexual and Reproductive Health Education | 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 Can AI Teach Sex Ed? A Systematic Review of the Use of Artificial Intelligence in Sexual and Reproductive Health Education Scarlett Bergam, Chloe Bergam, Brian Christopher Zanoni This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6289967/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 Background: Artificial intelligence (AI) is evolving and expanding at an unprecedented rate across healthcare and education. AI for sexual health education has the potential to reduce sexual health stigma, provide convenience for many populations of all genders, sexualities, and ages who were previously receiving insufficient or outdated information, and reduce the resources needed to provide this essential education. The aim of this systematic review is to assess the acceptability, feasibility, and impact of generative AI in sexual and reproductive health education. Methods : We searched PubMed, Web of Science, and Scopus in August 2024 combining artificial intelligence and sexual education search terms. We included experimental and observational studies of any analysis technique published between 01/01/2014-8/16/2024. Data was managed in Covidence. Screening and extraction utilized two non-expert reviewers. Quality assessment utilized the Mixed Methods Appraisal Tool and reporting adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist. Results: Initial database search yielded 4,044 records, 21 full-text articles meeting inclusion criteria. All were observational studies. Data included 146,990 individual participants (mean=7000, median=100) from eight countries. Five (23.8%) compared an AI chatbot to another format of sex education. Eighteen studies assessed acceptability, 12 studies assessed feasibility, and 13 studies assessed impact. Users of AI primarily seek factual information, find the chatbot's responses easy to understand, and appreciate the immediate responses compared to human responses. AI helps users exercise sexual rights, discuss sexual feelings/needs, and learn information about HIV and family planning. However, chatbot responses differ in tone and empathy than human responses and require long reading times. While chatbots are generally viewed as clinically safe and hold potential for providing accessible sexual health information, users show skepticism about their credibility for sensitive topics compared to human interactions. Conclusions: Usage of AI is surpassing high-quality evidence about its acceptability, feasibility, and impact. While initial studies show promise of AI chatbots for presenting sexual health information, high-quality, randomized studies with human participants and comparator groups are needed before AI can be trusted to successfully deliver such education. artificial intelligence sex education sex ed sexual health education reproductive health reproductive health education chatbot large language model generative AI AI Figures Figure 1 Introduction Sexual and reproductive health education is primarily delivered to adolescents in the school setting, and more recently through digital interventions and blended learning programs 1 . However, school-based sex education has been found to be insufficient in reducing risky health behaviors in teenagers while also excluding sexual minority youth in standardized curricula 2 . Outside of organized educational curricula, adolescents receive sex education by clinicians, caregivers, peers, and most commonly, the internet, although these sources of sex education are not regulated and pose the risk of providing misinformation and perpetuating stigma 3 . Artificial intelligence (AI) is expanding and evolving at an unprecedented rate across the healthcare and educational sectors, becoming an increasingly advanced and instrumental tool for educating patients, healthcare trainees, and medical providers alike 4 , 5 . One novel use of AI is to provide sexual and reproductive health education to the general public, including through formal interventions in clinical trials as well as through unregulated, informal use of free generative AI chatbots by individuals 6 . A 2021 commentary in The Lancet called for the use of artificial intelligence to enhance STI prevention and control, including for safer sex education 6 . We hypothesize that the usage of artificial intelligence-powered chatbots for sexual health education has the potential to reduce sexual health stigma, provide convenience for populations of all genders, sexualities, and ages who were previously receiving insufficient or outdated information, and reduce the labor needed to provide this education 6 . The role of artificial intelligence in sex education is a new and emerging field, which has been studied in recent years but has yet to be systematically reviewed. If AI is found to be a potentially successful strategy for sexual health education, this could have implications for transforming public health interventions and patient education in the modern era. Objectives The aim of this systematic review is to assess the impact, acceptability, and feasibility of generative AI in sexual and reproductive health education in the published literature. Methods Search Strategy and Selection Criteria We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement for the conduct of systematic reviews 7 . A review protocol was not prepared and this review was not registered with PROSPERO. On August 16th, 2024, we searched PubMed, Scopus, and Web of Science to avoid publication bias and include a wide range of journals within the fields of medicine, public health, social science, education, and technology. Our search string included terms related to artificial intelligence—("artificial intelligence"[tiab] OR "AI"[tiab] OR "machine learning"[tiab] OR "deep learning"[tiab] OR "chatbot"[tiab] OR "natural language processing"[tiab] OR "predictive modeling"[tiab] OR "generative AI"[tiab] OR "Artificial Intelligence"[Mesh])—as well as terms related to sexual education—("sex ed*" OR "sexual health" OR "reproductive education" OR "reproductive health" OR "family planning" OR "contraception education" OR "sexual health literacy" OR "Sex Education"[Mesh]). Detailed search strategy and selection criteria can be found in Appendix 2 . Original research articles were included if they: (a) involved human use of an artificial-intelligence powered-chatbot, (b) delivered information about sexual or reproductive health, including sexually transmitted infections, HIV, pregnancy prevention or fertility awareness, or sexuality, (c) assessed the acceptability, impact, and/or feasibility of the intervention, and (e) were published between January 1st, 2014 and August 16th, 2024. Exclusion criteria were as follows: (a) the content was general to other sectors of health or medicine or focused on diagnosis over education or (b) the article type was a commentary, editorial, review, or protocol lacking original data. We did not exclude on the basis of participant country, language, age, or gender, or whether or not the study contained a comparator group. Data management for this systematic review was facilitated using Covidence 8 , an online tool used for the management of systematic reviews to ensure an organized and repeatable workflow, from import of search results, screening, full-text review, and the management and storage of references. Screening and Extraction Screening We utilized a two stage screening process with two non-expert reviewers (SB and CB) independently conducting the title and abstract screening, followed by the full-text screening. Conflicts at each stage were resolved by discussion between reviewers while referencing a priori eligibility criteria. Data Extraction Data was extracted by one non-expert extractor (SB), with another non-expert (CB) serving as a verifier. Discordant results were resolved by discussion between the two reviewers and/or by consulting a senior researcher (BCZ). We extracted the following information from qualifying studies: study ID, title, last name of first author, country of study, publication type (gray vs. peer-reviewed literature), study design, aim of study, analysis type, start date, end date, population description, inclusion criteria, exclusion criteria, method of recruitment, total number of participants, format of intervention, topic of education, comparators, and outcomes (as detailed below). Outcomes We included studies that assessed either the acceptability, feasibility, and impact including both qualitative and quantitative findings, as interpreted by the reviewers based on the following a priori definitions. Acceptability was defined as participants' experiences interacting with artificial intelligence, including the perceived trustworthiness, tolerability, and satisfaction. Feasibility was defined as the ease of use of the chatbot, given the user interface and technological availability. Impact was defined as the ability of the AI intervention to change sexual health knowledge, attitudes, or behaviors. Data Synthesis Outcomes were qualitatively synthesized due to the variable outcome formats reported in the literature. Therefore, a meta-analysis with quantitative synthesis was not possible for this study. Data conversations were conducted between two authors to synthesize and analyse results. Microsoft Excel was used to tabulate and visually display results and syntheses. Quality Assessment To assess the quality of included studies, we used the Mixed Methods Appraisal Tool (MMAT), version 2018, to account for various study designs included in this analysis 9 . This tool has the ability to assess the risk of bias in qualitative, randomized quantitative, and non-randomized quantitative studies through appraisal questions verified for each study design. It consists of seven questions assessing the clarity of the research questions, focus of the data, appropriateness of the research approach, and substantiation of the findings from the data. Answers for each individual question included “Yes”, “No”, or “Can’t Tell”, with “Yes" scored as one point and “No” or “Can’t Tell” scored as 0 points. Quality and certainty of evidence was analyzed using Grading of Recommendations Assessment, Development and Evaluation (GRADE) criteria 10 , ranging from Very Low to High quality. Study quality was scored out of seven for each of the MMAT questions. Quality was considered “very low” if total GRADE score was 0–1, “low” if GRADE score was 2–3, “moderate” if GRADE score was 4–5, and “high” if GRADE score was 6–7. Results Our initial search yielded 4,044 records (651 on PubMed, 1,938 results on Scopus, and 1,455 results on Web of Science) on August 16th, 2024 as indicated in Fig. 1 . Results were uploaded to Covidence, where 723 duplicates were removed. Of the remaining 3,321 records, 3,213 titles and abstracts were removed due to irrelevance, leaving 108 articles for full-text eligibility review. Inter-rater reliability at this stage resulted in a Cohen’s kappa of 0.30 (moderate agreement). At the full-text review, 77 articles were excluded for wrong study design (n = 43), wrong intervention (n = 17), wrong indication (n = 13), and wrong patient population (n = 4). Excluded studies and their reason for exclusion can be found in Appendix 3. Twenty-one full-text articles were included in the review (Table 1). Among the 21 full-text articles included, 18 (85.7%) were peer-reviewed published manuscripts 11 – 28 and three (14.3%) were published conference abstracts 29 – 31 . Articles included 17 case studies 11–15,20−31 , one cohort study 16 , and three cross-sectional studies 17 – 19 . Most (61.9%) were published in 2023 and 2024. Seven (33.3%) undertook a mixed-methods analysis 11 , 13 , 14 , 16 , 18 , 20 , 28 , seven (33.3%) had primarily quantitative findings 12 , 15 , 17 , 24 , 26 , 27 , 29 , and seven (33.3%) had primarily qualitative findings 19 , 21 – 23 , 25 , 30 , 31 . Population The total population size of the studies was n = 146,990 individual participants (mean = 7000, median = 100) from eight countries, including seven (33.3%) from the United States 14 , 22 , 23 , 25 , 27 , 29 , 31 , five (23.8%) from the United Kingdom 12 , 17 – 19 , 30 , three (14.3%) from India 13 , 20 , 24 , , and three (14.3%) from Kenya 16 , 20 , 21 . Intervention Intervention content included thirteen studies with sex education that focused on general sexual health 11 , 13 – 18 , 20 , 21 , 26 – 28 , 30 , nine studies that focused on pregnancy prevention 11 , 14 , 20 , 22 , 24 , 26 , 27 , 29 , 31 , and nine studies that focused on sexually transmitted infection (including HIV) prevention and management 11 , 12 , 14 , 19 , 20 , 23 , 25 – 27 . Multiple interventions had more than one educational focus. Comparators Only five studies (23.8%) compared AI-generated sex education to another format of sex education 14 , 15 , 17 , 24 , 30 . Of these five, one compared chatbot responses to Google search results 14 , two compared to clinician-generated responses 17 , 30 , and two compared two AI chatbots to one another 15 , 24 . Quality Assessment Based on risk of bias analysis conducted with the MMAT and compiled with GRADE criteria, one study was considered “Very Low” quality 30 , three were “Low” quality 28 , 29 , 31 , five were “Medium” quality 15 – 18 , 24 , and 12 were “High” quality 11–14,19−23,25–27 . The mean MMAT score across study designs was 5.4/7. Full quality analysis can be found in Appendix 4 . Outcomes Outcomes of individual studies are summarized in Table 1 . Per reviewer categorization into a priori definitions of acceptability, feasibility, and impact; 18 studies (85.7%) assessed acceptability 11–14,16−21,23,25–31 , 12 studies (57.1%) assessed feasibility 11 , 13 – 16 , 19 , 20 , 23 , 25 , 28 – 30 , and 13 studies (61.9%) assessed impact 11 – 16 , 18 , 22 – 24 , 28 , 30 – 31 in study outcomes. Acceptability Positive attitudes towards AI chatbots including feeling that the SRH information provided was accurate, trustworthy, confidential, and reliable 11 – 13 , 29 and provided a confidential and judgement-free source of on-demand education that was accessible and informative 16 , 19 , 21 . Studies showed that users of AI chatbots would recommend them to others 11 , saying that it is a good use of time 11 and sparked further conversations about sexual health with trusted educators and healthcare providers 11 . Those who were more likely to find chatbots acceptable, according to one study 17 , were younger than 25 years old, Caucasian, and owned a smartphone. However, chatbots were found to be inclusive of users coming from diverse and underserved backgrounds, especially women and youth from rural backgrounds 13 . Two studies 26 , 29 revealed that performance expectancy—the belief that using a technology will be effective—impacted the acceptability of chatbots to provide SRH information. This theory was actualized in another study, when warning labels placed on AI-generated education on abortion negatively affected participant perceptions of the information 31 . Chatbots implemented in real-world settings, such as askNivi in Kenya and India 20 , SnehAI in India 13 , and Layla’s Got You in the United States 27 showed high total engagement across the population 13 , 20 , but with individual users utilizing the chatbot on two or fewer occasions 13 . Teenage and young adult users inquired about a wide variety of topics, including family planning, pregnancy, menstruation, and sex 20 , as well as acne, nocturnal ejaculation, sexually transmitted diseases, and pornography addiction 28 . In older populations, such as cancer survivors, chatbots were used to discuss sexual functioning, sexual response, body image, and intimacy 15 . However, other studies found acceptability scores as low as 40% 17 . Some felt that AI-generated SRH information was difficult to comprehend, took longer to read, and were less accurate with current evidence, especially when compared to Google Search results 14 . When compared to clinician generated-responses, they were clearly identified as AI generated by participants 30 . While it was seen as acceptable for general sexual health advice and “signposting”—referring to a healthcare provider rather than giving medical advice—it was not perceived as an acceptable method of diagnosis or emotional support 18 , 19 . Participants in one study 25 perceived interacting with a bot as a form of neglect compared to interacting with a human, due to lack of empathy. Feasibility Users utilized AI chatbots for multiple formats of SRH education—including as a creator of factual information about symptoms and sexual health concepts, a source of personal advice, and as a place to request a diagnosis 20 . From the perspective of research staff, having study participants utilize SRH chatbots was seen as a way to reduce their workload and set boundaries outside of work hours 25 . Participants even perceived chatbots as “friends”, seeking out their advice when trusted individuals were not around to provide answers 28 . Although AI-generated education was praised for being generated quickly 30 , it required longer reading times than Google Search results 14 or clinician-generated responses 30 . Many considered chatbots inferior to healthcare professionals or internet search results due to distrust and low perceived accuracy 19 . Using the Flesh-Kincaid score, a verified metric for readability of the written English language, one study showed that chatbot-based SRH information was written at the college graduate or professional level 14 . Impact In one study of clinicians, 22% perceived chatbot responses as effective and 24% perceived them as ineffective for SRH advice 18 . From the participant perspective, the impact of chatbot-powered SRH education included participant-perceived increases in HIV and family planning knowledge 11 , positive impacts on intentions to get tested for STIs 12 , and subjective empowerment of young people to advocate for their sexual feelings and rights 16 . Chatbots were found to satisfy most users’ queries, often dispelling misconceptions and providing accurate education 28 , and correctly counseling users to seek professional medical advice 31 . However, the impact of AI-based education was hindered by its occasional factual inaccuracy, excessive information, and poor organization 14 , 22 . Not all AI chatbots are equally impactful—one study found superior accuracy and completeness of answers provided by ChatGPT compared to Google Bard AI 24 . Furthermore, AI-generated answers are often better at answering some SRH-related queries than others. One study found that information provided about PrEP and STI by a chatbot scored higher than clinician-generated responses, but its information about contraception was seen as inferior to human-created education. Discussion This systematic review synthesizes the available literature on the use of artificial intelligence (AI) in sexual and reproductive health education, assessing its acceptability, feasibility, and impact, highlighting both the promises and the challenges associated with AI-driven sex education interventions. The acceptability of AI-based interventions in the healthcare space is based on user factors such as trust, literacy, and receptiveness; system factors such as workflow integration; and social factors such as ethics 36 . Overall, AI-based sexual health education interventions appear to be widely accepted across diverse populations for providing immediate, accessible, and often judgment-free responses, making them particularly attractive to individuals seeking confidential sexual health information 32 . The studies included in this review suggest that users appreciate the convenience, ease of use, and anonymity provided by AI chatbots, particularly for topics that may be stigmatized in traditional educational settings. However, concerns remain about the credibility, emotional disconnect, and readability that AI-based interventions can provide compared to trained sexual health educators or healthcare professionals 33 . The feasibility of AI-chatbots in public health interventions revolves around its ability to provide a non-judgemental space to discuss sensitive health information, and the potential for scalability 37 . The feasibility of AI-driven sex education was supported by several studies. Chatbots demonstrated technical functionality, the ability to tailor responses to diverse user populations, ease of implementation fulfilling the role of quickly generating information, and the potential for reducing clinician workload. The ability of AI to deliver personalized educational content and respond to user queries in real-time enhances its potential as a scalable public health intervention. However, feasibility was limited in some studies by the variations in chatbot comprehension, accuracy, and reading times. With regard to impact, outcomes were mixed. Some studies included in this review indicate that AI chatbots can often provide factual sexual health information, and subjectively increased users’ confidence discussing sexual feelings and needs, and created positive behavioral intentions. However, other studies revealed that some AI chatbots functioned better than other chatbots, and certain topics were better taught by artificial intelligence than other topics. Ultimately, the lack of rigorous randomized experimental designs limits the ability to draw definitive conclusions about AI’s efficacy, or even its superiority or inferiority compared to human-led sex education programs. Several key recommendations emerge from this review. First, future research should prioritize the development and evaluation of AI-driven sex education interventions using high-quality study designs, including randomized controlled trials (RCTs) with appropriate control groups. Additionally, AI-driven interventions should be integrated into broader educational and healthcare systems to ensure that they complement, rather than replace, human-led education and counseling 35 . Finally, further investigation is needed to assess the long-term efficacy of AI-based sex education on behavioral outcomes, including safer sex practices, STI prevention, and contraceptive use. A major strength of this systematic review is its comprehensive evaluation of AI-driven sexual health education across multiple dimensions, including acceptability, feasibility, and impact. The inclusion of studies from diverse geographic regions and gray literature enhances the generalizability of the findings while reducing publication bias. Additionally, the use of a standardized quality assessment tool (MMAT) strengthens the validity of this review 9 . Several limitations must be acknowledged. First, the lack of RCTs limit the ability to establish causal relationships between AI interventions and improved sexual health knowledge or behaviors, with all 21 studies lacking results that show clear efficacy. In addition, heterogeneity between study design and publication type intrinsically reduces the comparability of findings, preventing our ability to complete a meta-analysis. To standardize quality analysis, we utilized the Mixed-Methods Appraisal Tool to assess all studies on the basis of their individual study designs. Further, moderate inter-rater reliability in the screening round was resolved with additional reviewer training and discussion on a priori eligibility criteria. Additionally, large language models are trained by a limited scope and sometimes false or biased information, which has the potential to lead to knowledge gaps or misinformation in fields such as abortion and LGBTQ + health. Finally, the rapid evolution of AI technology means that some of the findings may become outdated as newer, more sophisticated AI models emerge. Conclusions While AI has demonstrated promise in delivering accessible and user-friendly sexual health education, the current evidence-base remains limited. AI chatbots and virtual assistants have the potential to complement traditional sex education methods, particularly in addressing stigma and providing real-time information. However, significant gaps remain in assessing their research-based efficacy and real-world effectiveness relative to human-led approaches. Future research should focus on rigorously evaluating AI’s role in sexual health education to ensure that it is both a credible and effective tool for modern public health advancement. Abbreviations Artificial intelligence (AI) Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) Mixed Methods Appraisal Tool (MMAT) Randomized Control Trial (RCT) Declarations Ethics approval and consent to participate: Not applicable Consent for publication: Not applicable Availability of data and materials: Template data collection forms and data extracted from the included studies are available by request Competing interests: The authors report there are no competing interests to declare. Funding: This work was supported by the 2024 Lazarus Family Scholarship at George Washington University. Authors' Contributions: Study conceptualization was led by SB under the guidance of BCZ. SB and CB participated in data search, screening, data extraction, and quality assessment. SB led data synthesis and preparation of the manuscript, tables, and figures. All authors edited and approved the final manuscript for submission. Acknowledgements: Thank you to Ms. Laura Abate and Mr. Tom Harrod, librarians at the George Washington University School of Medicine and Health Sciences, for their guidance on systematic review methods. References Lameiras-Fernández M, Martínez-Román R, Carrera-Fernández MV, Rodríguez-Castro Y. Sex education in the spotlight: what Is working? Systematic review. Int J Environ Res Public Health. 2021;18(5):2555. doi: 10.3390/ijerph18052555 . PMID: 33806507; PMCID: PMC7967369. Bishop MD, Mallory AB, Gessner M, Frost DM, Russell ST. School-based sexuality education experiences across three generations of sexual minority people. J Sex Res. 2021 May-Jun;58(5):648–658. doi: 10.1080/00224499.2020.1767024. Epub 2020 Jun 2. PMID: 32486928; PMCID: PMC7708407. Gray BJ, Jones AT, Couzens Z, Sagar T, Jones D. University students' behaviours towards accessing sexual health information and treatment. Int J STD AIDS. 2019;30(7):671–679. doi: 10.1177/0956462419828866. Epub 2019 Apr 8. PMID: 30961468. Barbour AB, Frush JM, Gatta LA, McManigle WC, Keah NM, Bejarano-Pineda L, Guerrero EM. Artificial intelligence in health care: insights from an educational forum. J Med Educ Curric Dev. 2020;6:2382120519889348. doi: 10.1177/2382120519889348 . PMID: 32064356; PMCID: PMC6993147. Dave M, Patel N. Artificial intelligence in healthcare and education. Br Dent J. 2023;234, 761–764. https://doi.org/10.1038/s41415-023-5845-2 Young SD, Crowley JS, Vermund SH. Artificial intelligence and sexual health in the USA. Lancet Digit Health. 2021;3(8):e467-e468. doi: 10.1016/S2589-7500(21)00117-5. PMID: 34325852; PMCID: PMC10767714. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. doi: 10.1136/bmj.n71 . Covidence systematic review software, Veritas Health Innovation, Melbourne, Australia. Available at www.covidence.org Hong QN, Pluye P, Fàbregues S, Bartlett G, Boardman F, Cargo M, et al. Mixed methods appraisal tool (MMAT) version 2018 user guide. McGill Department of Family Medicine. 2018 Aug 1. Centers for Disease Control and Prevention (CDC). GRADE criteria for determining certainty of evidence [Internet]. Atlanta, GA: CDC; [cited 2025 Feb 27]. Available from: https://www.cdc.gov/acip-grade-handbook/hcp/chapter-7-grade-criteria-determining-certainty-of-evidence/index.html . Yam EA, Namukonda E, McClair T, Souidi S, Chelwa N, Muntalima N, et al. Developing and testing a chatbot to integrate HIV education into family planning clinic waiting areas in Lusaka, Zambia. Glob Health Sci Pract. 2022;10(5). doi: 10.9745/GHSP-D-21-00721 . Nadarzynski T, Knights N, Husbands D, Graham CA, Llewellyn CD, Buchanan T, et al. The impact of Chatbot-Assisted Self-Assessment (CASA) on intentions for sexual health screening in people from minoritised ethnic groups at risk of sexually transmitted infections. Sex Health. 2024;21. doi: 10.1071/SH24058 . Wang H, Gupta S, Singhal A, Muttreja P, Singh S, Sharma P, et al. An artificial intelligence chatbot for young people's sexual and reproductive health in India (SnehAI): instrumental case study. J Med Internet Res. 2022;24(1):e29969. doi: 10.2196/29969 . Burns C, Bakaj A, Berishaj A, Hristidis V, Deak P, Equils O. Use of generative AI for improving health literacy in reproductive health: case study. JMIR Form Res. 2024;8:e59434. doi: 10.2196/59434 . Hanai A, Ishikawa T, Kawauchi S, Iida Y, Kawakami E. Generative artificial intelligence and non-pharmacological bias: an experimental study on cancer patient sexual health communications. BMJ Health Care Inform. 2024;31(1). doi: 10.1136/bmjhci-2023-100924 . Njogu J, Jaworski G, Oduor C, Chea A, Malmqvist A, Rothschild CW. Assessing acceptability and effectiveness of a pleasure-oriented sexual and reproductive health chatbot in Kenya: an exploratory mixed-methods study. Sex Reprod Health Matters. 2023;31(4):2269008. doi: 10.1080/26410397.2023.2269008 . Nadarzynski T, Bayley J, Llewellyn C, Kidsley S, Graham CA. Acceptability of artificial intelligence (AI)-enabled chatbots, video consultations and live webchats as online platforms for sexual health advice. BMJ Sex Reprod Health. 2020;46(3):210–217. doi: 10.1136/bmjsrh-2018-200271 . Nadarzynski T, Lunt A, Knights N, Bayley J, Llewellyn C. "But can chatbots understand sex?" Attitudes towards artificial intelligence chatbots amongst sexual and reproductive health professionals: an exploratory mixed-methods study. Int J STD AIDS. 2023;34(11):809–816. doi: 10.1177/09564624231180777 . Nadarzynski T, Puentes V, Pawlak I, Mendes T, Montgomery I, Bayley J, et al. Barriers and facilitators to engagement with artificial intelligence (AI)-based chatbots for sexual and reproductive health advice: a qualitative analysis. Sex Health. 2021;18(5):385–393. doi: 10.1071/SH21123 . Green EP, Whitcomb A, Kahumbura C, Rosen JG, Goyal S, Achieng D, et al. "What is the best method of family planning for me?": a text mining analysis of messages between users and agents of a digital health service in Kenya. Gates Open Res. 2019;3:1475. doi: 10.12688/gatesopenres.12999.1 . Ngönjiri A, Memiah P, Kimathi R, Wagner FA, Ikahu A, Omanga E, et al. Utilizing user preferences in designing the AGILE (Accelerating Access to Gender-Based Violence Information and Services Leveraging on Technology Enhanced) chatbot. Int J Environ Res Public Health. 2023;20(21). doi: 10.3390/ijerph20217018 . McMahon HV, McMahon BD. Automating untruths: ChatGPT, self-managed medication abortion, and the threat of misinformation in a post-Roe world. Front Digit Health. 2024;6:1287186. doi: 10.3389/fdgth.2024.1287186 . Chandler RD, Warner S, Aidoo-Frimpong G, Wells J. "What Did You Say, ChatGPT?" The use of AI in Black women's HIV self-education: an inductive qualitative data analysis. J Assoc Nurses AIDS Care. 2024 May-Jun;35(3):294–302. doi: 10.1097/JNC.0000000000000468 . Mediboina A, Badam RK, Chodavarapu S. Assessing the accuracy of information on medication abortion: a comparative analysis of ChatGPT and Google Bard AI. Cureus. 2024;16(1):e51544. doi: 10.7759/cureus.51544 . Comulada WS, Rezai R, Sumstine S, Flores DD, Kerin T, Ocasio MA, et al. A necessary conversation to develop chatbots for HIV studies: qualitative findings from research staff, community advisory board members, and study participants. AIDS Care - Psychological and Socio-Medical Aspects of AIDS/HIV. 2024;36(4):463–471. doi: 10.1080/09540121.2023.2216926 . Liew TW, Tan S-M, Yoo NE, Gan CL, Lee YY. Let's talk about Sex!: AI and relational factors in the adoption of a chatbot conveying sexual and reproductive health information. Comput Hum Behav Rep. 2023;11:100323. doi: 10.1016/j.chbr.2023.100323 . Bonnevie E, Lloyd TD, Rosenberg SD, Williams K, Goldbarg J, Smyser J. Layla's Got You: Developing a tailored contraception chatbot for Black and Hispanic young women. Health Educ J. 2021;80(4):413–424. doi: 10.1177/0017896920981122 . Rahman R, Rahman MR, Tripto NI. Adolescentbot: Understanding opportunities for chatbots in combating adolescent sexual and reproductive health problems in Bangladesh. In: Conference on Human Factors in Computing Systems - Proceedings; 2021. doi: 10.1145/3411764.3445694 . Hussain SA, Ogundimu F, Bhattarai S. Mobile phone-based chatbot for family planning and contraceptive information. Lecture Notes in Computer Science; 2019. p. 342–352. doi: 10.1007/978-3-030-22219-2_26 . Taylor R, Messham R, Okafor C, Pink E, Portas L, Gees H. Can sexual health clinicians be replaced by robots? Utility of artificial intelligence platforms to give sexual health advice. Sex Transm Infect. 2023;99(Suppl 1):A8-A8. doi: 10.1136/sextrans-BASHH-2023.16 . Sharevski F, Loop JV, Jachim P, Devine A, Pieroni E. Talking abortion (mis)information with ChatGPT on TikTok. In: Proceedings – 8th IEEE European Symposium on Security and Privacy Workshops, Euro S and PW 2023; 2023; 594–608. doi: 10.1109/EuroSPW59978.2023.00071 . Mehta N, Gupta S, Kularathne Y. The role and impact of artificial intelligence in addressing sexually transmitted infections, nonvenereal genital diseases, sexual health, and wellness. Indian Dermatol Online J. 2023;14(6):793–798. doi: 10.4103/idoj.idoj_426_23 . PMID: 38099049; PMCID: PMC10718125. Kerasidou A. Artificial intelligence and the ongoing need for empathy, compassion and trust in healthcare. Bull World Health Organ. 2020;98(4):245–250. doi: 10.2471/BLT.19.237198 . Epub 2020 Jan 27. PMID: 32284647; PMCID: PMC7133472. Zhai C, Wibowo S, Li LD. Evaluating the AI dialogue system's intercultural, humorous, and empathetic dimensions in English language learning: A case study. Comput Educ Artif Intell. 2024;7:100262. Maleki Varnosfaderani S, Forouzanfar M. The role of AI in hospitals and clinics: transforming healthcare in the 21st Century. Bioengineering (Basel). 2024;11(4):337. doi: 10.3390/bioengineering11040337 . PMID: 38671759; PMCID: PMC11047988. Hua D, Petrina N, Young N, Cho JG, Poon SK. Understanding the factors influencing acceptability of AI in medical imaging domains among healthcare professionals: A scoping review. Artif Intell Med. 2024;147:102698. doi: 10.1016/j.artmed.2023.102698 . Epub 2023 Nov 9. PMID: 38184343. Aggarwal A, Tam CC, Wu D, Li X, Qiao S. Artificial Intelligence-Based Chatbots for Promoting Health Behavioral Changes: Systematic Review. J Med Internet Res. 2023;25:e40789. doi: 10.2196/40789 . PMID: 36826990; PMCID: PMC10007007. Tables Table 1 is available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files Table1.xlsx Table 1: Results of Included Studies APPENDIX1.docx Appendix 1: Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Checklist Appendix2.docx Appendix 2: Databases and Search Strategy Appendix3.docx Appendix 3: Excluded Studies and Reason for Exclusion Appendix4.docx Appendix 4: Complete Quality Assessment by Study 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-6289967","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":433281927,"identity":"34697a02-8a3c-4d9d-a84d-92f7effe683d","order_by":0,"name":"Scarlett Bergam","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBElEQVRIiWNgGAWjYLCCB0BsAGHaADFj4wGCWhIYDCSgWtJAWhpI0nIYTOLVIt9++NmHhIo/debsvY9f/Mw5b7e2/TDQlhqbaFxaDM6kGc9IOGMgYdlz3Myyd9vt5G1nEoFajqXlNuDSIsFgzJDYBnTYjTQ2A16gFrMDQC2MDYdxapGfwf6ZIfEfUMv9Z2yGf7edSzY7/xC/FoYbPEBbGkC2sDE/5t12wM7sBgFbDM7kFDMkHDOW3NmTxsYsuy05wewG0JYEPH6Rbz++meFDjRy/Ofsx5o9vt9nZm51Pf/jgQ40NbochATYJIJEIVplAhHIQYP4AJOyJVDwKRsEoGAUjCAAAihhis7JG13cAAAAASUVORK5CYII=","orcid":"","institution":"George Washington University","correspondingAuthor":true,"prefix":"","firstName":"Scarlett","middleName":"","lastName":"Bergam","suffix":""},{"id":433281928,"identity":"d4d2b954-fc41-4f7f-85ee-f9c94a0b95ee","order_by":1,"name":"Chloe Bergam","email":"","orcid":"","institution":"Tufts University","correspondingAuthor":false,"prefix":"","firstName":"Chloe","middleName":"","lastName":"Bergam","suffix":""},{"id":433281929,"identity":"3193c9e1-ec82-4114-ba07-528deba4499a","order_by":2,"name":"Brian Christopher Zanoni","email":"","orcid":"","institution":"Emory University School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Brian","middleName":"Christopher","lastName":"Zanoni","suffix":""}],"badges":[],"createdAt":"2025-03-23 19:38:09","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6289967/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6289967/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":79262733,"identity":"fec50af7-046c-49fb-82f3-b4a5fe501a4c","added_by":"auto","created_at":"2025-03-26 09:47:22","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":293434,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePreferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFlow Diagram\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6289967/v1/2415421fb82cacebcaa0bf21.png"},{"id":84870995,"identity":"1a621731-fbf7-476a-af3e-8245bf3ae74d","added_by":"auto","created_at":"2025-06-18 09:02:15","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":831017,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6289967/v1/0497a6b5-4a11-4333-9570-05735b86d045.pdf"},{"id":79262725,"identity":"a9b83548-3546-409c-ab56-ef55ddd7d586","added_by":"auto","created_at":"2025-03-26 09:47:22","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":45805,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTable 1: Results of Included Studies\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"Table1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6289967/v1/499689f8e52e1c45a2b18325.xlsx"},{"id":79264672,"identity":"dd09b04f-39be-4c17-b54e-4fb0e91ed396","added_by":"auto","created_at":"2025-03-26 09:55:22","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":17763,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAppendix 1: Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Checklist\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"APPENDIX1.docx","url":"https://assets-eu.researchsquare.com/files/rs-6289967/v1/cf34f6537c3c36b50c378b94.docx"},{"id":79262728,"identity":"0462d540-06dd-4c98-a349-4c6c75cb96c7","added_by":"auto","created_at":"2025-03-26 09:47:22","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":15085,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAppendix 2: Databases and Search Strategy\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"Appendix2.docx","url":"https://assets-eu.researchsquare.com/files/rs-6289967/v1/a272ed65b076c563496d7125.docx"},{"id":79264673,"identity":"ce2fce47-955e-4e67-9643-c6562ced6673","added_by":"auto","created_at":"2025-03-26 09:55:22","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":28289,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAppendix 3: Excluded Studies and Reason for Exclusion\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"Appendix3.docx","url":"https://assets-eu.researchsquare.com/files/rs-6289967/v1/ba396a0ae392545301cb8eb6.docx"},{"id":79262732,"identity":"f9d91668-39e5-4a03-97dc-9ff0f8338362","added_by":"auto","created_at":"2025-03-26 09:47:22","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":21609,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAppendix 4: Complete Quality Assessment by Study\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"Appendix4.docx","url":"https://assets-eu.researchsquare.com/files/rs-6289967/v1/fa3c27f0bd2cb23e7ca3b192.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Can AI Teach Sex Ed? A Systematic Review of the Use of Artificial Intelligence in Sexual and Reproductive Health Education","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSexual and reproductive health education is primarily delivered to adolescents in the school setting, and more recently through digital interventions and blended learning programs\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. However, school-based sex education has been found to be insufficient in reducing risky health behaviors in teenagers while also excluding sexual minority youth in standardized curricula\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Outside of organized educational curricula, adolescents receive sex education by clinicians, caregivers, peers, and most commonly, the internet, although these sources of sex education are not regulated and pose the risk of providing misinformation and perpetuating stigma\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eArtificial intelligence (AI) is expanding and evolving at an unprecedented rate across the healthcare and educational sectors, becoming an increasingly advanced and instrumental tool for educating patients, healthcare trainees, and medical providers alike\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. One novel use of AI is to provide sexual and reproductive health education to the general public, including through formal interventions in clinical trials as well as through unregulated, informal use of free generative AI chatbots by individuals\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. A 2021 commentary in \u003cem\u003eThe Lancet\u003c/em\u003e called for the use of artificial intelligence to enhance STI prevention and control, including for safer sex education\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eWe hypothesize that the usage of artificial intelligence-powered chatbots for sexual health education has the potential to reduce sexual health stigma, provide convenience for populations of all genders, sexualities, and ages who were previously receiving insufficient or outdated information, and reduce the labor needed to provide this education\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. The role of artificial intelligence in sex education is a new and emerging field, which has been studied in recent years but has yet to be systematically reviewed. If AI is found to be a potentially successful strategy for sexual health education, this could have implications for transforming public health interventions and patient education in the modern era.\u003c/p\u003e\u003ch2\u003eObjectives\u003c/h2\u003e\u003cp\u003eThe aim of this systematic review is to assess the impact, acceptability, and feasibility of generative AI in sexual and reproductive health education in the published literature.\u003c/p\u003e"},{"header":"Methods","content":" \u003cdiv id=\"Sec4\" class=\"Section3\"\u003e \u003ch2\u003eSearch Strategy and Selection Criteria\u003c/h2\u003e \u003cp\u003eWe followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement for the conduct of systematic reviews\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. A review protocol was not prepared and this review was not registered with PROSPERO.\u003c/p\u003e \u003cp\u003eOn August 16th, 2024, we searched PubMed, Scopus, and Web of Science to avoid publication bias and include a wide range of journals within the fields of medicine, public health, social science, education, and technology. Our search string included terms related to artificial intelligence\u0026mdash;(\"artificial intelligence\"[tiab] OR \"AI\"[tiab] OR \"machine learning\"[tiab] OR \"deep learning\"[tiab] OR \"chatbot\"[tiab] OR \"natural language processing\"[tiab] OR \"predictive modeling\"[tiab] OR \"generative AI\"[tiab] OR \"Artificial Intelligence\"[Mesh])\u0026mdash;as well as terms related to sexual education\u0026mdash;(\"sex ed*\" OR \"sexual health\" OR \"reproductive education\" OR \"reproductive health\" OR \"family planning\" OR \"contraception education\" OR \"sexual health literacy\" OR \"Sex Education\"[Mesh]). Detailed search strategy and selection criteria can be found in \u003cb\u003eAppendix 2\u003c/b\u003e.\u003c/p\u003e \u003cp\u003eOriginal research articles were included if they: (a) involved human use of an artificial-intelligence powered-chatbot, (b) delivered information about sexual or reproductive health, including sexually transmitted infections, HIV, pregnancy prevention or fertility awareness, or sexuality, (c) assessed the acceptability, impact, and/or feasibility of the intervention, and (e) were published between January 1st, 2014 and August 16th, 2024. Exclusion criteria were as follows: (a) the content was general to other sectors of health or medicine or focused on diagnosis over education or (b) the article type was a commentary, editorial, review, or protocol lacking original data. We did not exclude on the basis of participant country, language, age, or gender, or whether or not the study contained a comparator group.\u003c/p\u003e \u003cp\u003eData management for this systematic review was facilitated using Covidence\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e, an online tool used for the management of systematic reviews to ensure an organized and repeatable workflow, from import of search results, screening, full-text review, and the management and storage of references.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e\n\u003ch3\u003eScreening and Extraction\u003c/h3\u003e\n\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eScreening\u003c/h2\u003e \u003cp\u003e We utilized a two stage screening process with two non-expert reviewers (SB and CB) independently conducting the title and abstract screening, followed by the full-text screening. Conflicts at each stage were resolved by discussion between reviewers while referencing a priori eligibility criteria.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eData Extraction\u003c/h3\u003e\n\u003cp\u003eData was extracted by one non-expert extractor (SB), with another non-expert (CB) serving as a verifier. Discordant results were resolved by discussion between the two reviewers and/or by consulting a senior researcher (BCZ). We extracted the following information from qualifying studies: study ID, title, last name of first author, country of study, publication type (gray vs. peer-reviewed literature), study design, aim of study, analysis type, start date, end date, population description, inclusion criteria, exclusion criteria, method of recruitment, total number of participants, format of intervention, topic of education, comparators, and outcomes (as detailed below).\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eOutcomes\u003c/h2\u003e \u003cp\u003eWe included studies that assessed either the acceptability, feasibility, and impact including both qualitative and quantitative findings, as interpreted by the reviewers based on the following a priori definitions. Acceptability was defined as participants' experiences interacting with artificial intelligence, including the perceived trustworthiness, tolerability, and satisfaction. Feasibility was defined as the ease of use of the chatbot, given the user interface and technological availability. Impact was defined as the ability of the AI intervention to change sexual health knowledge, attitudes, or behaviors.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eData Synthesis\u003c/h3\u003e\n\u003cp\u003eOutcomes were qualitatively synthesized due to the variable outcome formats reported in the literature. Therefore, a meta-analysis with quantitative synthesis was not possible for this study. Data conversations were conducted between two authors to synthesize and analyse results. Microsoft Excel was used to tabulate and visually display results and syntheses.\u003c/p\u003e\n\u003ch3\u003eQuality Assessment\u003c/h3\u003e\n\u003cp\u003eTo assess the quality of included studies, we used the Mixed Methods Appraisal Tool (MMAT), version 2018, to account for various study designs included in this analysis\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. This tool has the ability to assess the risk of bias in qualitative, randomized quantitative, and non-randomized quantitative studies through appraisal questions verified for each study design. It consists of seven questions assessing the clarity of the research questions, focus of the data, appropriateness of the research approach, and substantiation of the findings from the data. Answers for each individual question included \u0026ldquo;Yes\u0026rdquo;, \u0026ldquo;No\u0026rdquo;, or \u0026ldquo;Can\u0026rsquo;t Tell\u0026rdquo;, with \u0026ldquo;Yes\" scored as one point and \u0026ldquo;No\u0026rdquo; or \u0026ldquo;Can\u0026rsquo;t Tell\u0026rdquo; scored as 0 points. Quality and certainty of evidence was analyzed using Grading of Recommendations Assessment, Development and Evaluation (GRADE) criteria\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e, ranging from Very Low to High quality. Study quality was scored out of seven for each of the MMAT questions. Quality was considered \u0026ldquo;very low\u0026rdquo; if total GRADE score was 0\u0026ndash;1, \u0026ldquo;low\u0026rdquo; if GRADE score was 2\u0026ndash;3, \u0026ldquo;moderate\u0026rdquo; if GRADE score was 4\u0026ndash;5, and \u0026ldquo;high\u0026rdquo; if GRADE score was 6\u0026ndash;7.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eOur initial search yielded 4,044 records (651 on PubMed, 1,938 results on Scopus, and 1,455 results on Web of Science) on August 16th, 2024 as indicated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Results were uploaded to Covidence, where 723 duplicates were removed. Of the remaining 3,321 records, 3,213 titles and abstracts were removed due to irrelevance, leaving 108 articles for full-text eligibility review. Inter-rater reliability at this stage resulted in a Cohen\u0026rsquo;s kappa of 0.30 (moderate agreement). At the full-text review, 77 articles were excluded for wrong study design (n\u0026thinsp;=\u0026thinsp;43), wrong intervention (n\u0026thinsp;=\u0026thinsp;17), wrong indication (n\u0026thinsp;=\u0026thinsp;13), and wrong patient population (n\u0026thinsp;=\u0026thinsp;4). Excluded studies and their reason for exclusion can be found in \u003cb\u003eAppendix 3.\u003c/b\u003e Twenty-one full-text articles were included in the review \u003cb\u003e(Table\u0026nbsp;1).\u003c/b\u003e\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAmong the 21 full-text articles included, 18 (85.7%) were peer-reviewed published manuscripts\u003csup\u003e\u003cspan additionalcitationids=\"CR12 CR13 CR14 CR15 CR16 CR17 CR18 CR19 CR20 CR21 CR22 CR23 CR24 CR25 CR26 CR27\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e and three (14.3%) were published conference abstracts\u003csup\u003e\u003cspan additionalcitationids=\"CR30\" citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Articles included 17 case studies\u003csup\u003e11\u0026ndash;15,20\u0026minus;31\u003c/sup\u003e, one cohort study\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e, and three cross-sectional studies\u003csup\u003e\u003cspan additionalcitationids=\"CR18\" citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Most (61.9%) were published in 2023 and 2024. Seven (33.3%) undertook a mixed-methods analysis\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e, seven (33.3%) had primarily quantitative findings\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e, and seven (33.3%) had primarily qualitative findings\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan additionalcitationids=\"CR22\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003ePopulation\u003c/h2\u003e \u003cp\u003eThe total population size of the studies was n\u0026thinsp;=\u0026thinsp;146,990 individual participants (mean\u0026thinsp;=\u0026thinsp;7000, median\u0026thinsp;=\u0026thinsp;100) from eight countries, including seven (33.3%) from the United States\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, five (23.8%) from the United Kingdom\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan additionalcitationids=\"CR18\" citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, three (14.3%) from India\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003c/sup\u003e, and three (14.3%) from Kenya\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eIntervention\u003c/h2\u003e \u003cp\u003eIntervention content included thirteen studies with sex education that focused on general sexual health\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan additionalcitationids=\"CR14 CR15 CR16 CR17\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan additionalcitationids=\"CR27\" citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, nine studies that focused on pregnancy prevention\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, and nine studies that focused on sexually transmitted infection (including HIV) prevention and management\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan additionalcitationids=\"CR26\" citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Multiple interventions had more than one educational focus.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eComparators\u003c/h2\u003e \u003cp\u003eOnly five studies (23.8%) compared AI-generated sex education to another format of sex education\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. Of these five, one compared chatbot responses to Google search results\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, two compared to clinician-generated responses\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, and two compared two AI chatbots to one another\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eQuality Assessment\u003c/h2\u003e \u003cp\u003eBased on risk of bias analysis conducted with the MMAT and compiled with GRADE criteria, one study was considered \u0026ldquo;Very Low\u0026rdquo; quality\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, three were \u0026ldquo;Low\u0026rdquo; quality\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, five were \u0026ldquo;Medium\u0026rdquo; quality\u003csup\u003e\u003cspan additionalcitationids=\"CR16 CR17\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e, and 12 were \u0026ldquo;High\u0026rdquo; quality\u003csup\u003e11\u0026ndash;14,19\u0026minus;23,25\u0026ndash;27\u003c/sup\u003e. The mean MMAT score across study designs was 5.4/7. Full quality analysis can be found in \u003cb\u003eAppendix 4\u003c/b\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eOutcomes\u003c/h2\u003e \u003cp\u003eOutcomes of individual studies are summarized in \u003cb\u003eTable\u0026nbsp;1\u003c/b\u003e. Per reviewer categorization into a priori definitions of acceptability, feasibility, and impact; 18 studies (85.7%) assessed acceptability\u003csup\u003e11\u0026ndash;14,16\u0026minus;21,23,25\u0026ndash;31\u003c/sup\u003e, 12 studies (57.1%) assessed feasibility\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan additionalcitationids=\"CR14 CR15\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan additionalcitationids=\"CR29\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, and 13 studies (61.9%) assessed impact\u003csup\u003e\u003cspan additionalcitationids=\"CR12 CR13 CR14 CR15\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan additionalcitationids=\"CR23\" citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e in study outcomes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eAcceptability\u003c/h2\u003e \u003cp\u003ePositive attitudes towards AI chatbots including feeling that the SRH information provided was accurate, trustworthy, confidential, and reliable\u003csup\u003e\u003cspan additionalcitationids=\"CR12\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e and provided a confidential and judgement-free source of on-demand education that was accessible and informative\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Studies showed that users of AI chatbots would recommend them to others\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e, saying that it is a good use of time\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e and sparked further conversations about sexual health with trusted educators and healthcare providers\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Those who were more likely to find chatbots acceptable, according to one study\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, were younger than 25 years old, Caucasian, and owned a smartphone. However, chatbots were found to be inclusive of users coming from diverse and underserved backgrounds, especially women and youth from rural backgrounds\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Two studies\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e revealed that performance expectancy\u0026mdash;the belief that using a technology will be effective\u0026mdash;impacted the acceptability of chatbots to provide SRH information. This theory was actualized in another study, when warning labels placed on AI-generated education on abortion negatively affected participant perceptions of the information\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eChatbots implemented in real-world settings, such as \u003cem\u003easkNivi\u003c/em\u003e in Kenya and India\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, \u003cem\u003eSnehAI\u003c/em\u003e in India\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e, and \u003cem\u003eLayla\u0026rsquo;s Got You\u003c/em\u003e in the United States\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e showed high total engagement across the population\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, but with individual users utilizing the chatbot on two or fewer occasions\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Teenage and young adult users inquired about a wide variety of topics, including family planning, pregnancy, menstruation, and sex\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, as well as acne, nocturnal ejaculation, sexually transmitted diseases, and pornography addiction\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. In older populations, such as cancer survivors, chatbots were used to discuss sexual functioning, sexual response, body image, and intimacy\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHowever, other studies found acceptability scores as low as 40%\u003csup\u003e17\u003c/sup\u003e. Some felt that AI-generated SRH information was difficult to comprehend, took longer to read, and were less accurate with current evidence, especially when compared to Google Search results\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. When compared to clinician generated-responses, they were clearly identified as AI generated by participants\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. While it was seen as acceptable for general sexual health advice and \u0026ldquo;signposting\u0026rdquo;\u0026mdash;referring to a healthcare provider rather than giving medical advice\u0026mdash;it was not perceived as an acceptable method of diagnosis or emotional support\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Participants in one study\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e perceived interacting with a bot as a form of neglect compared to interacting with a human, due to lack of empathy.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eFeasibility\u003c/h2\u003e \u003cp\u003eUsers utilized AI chatbots for multiple formats of SRH education\u0026mdash;including as a creator of factual information about symptoms and sexual health concepts, a source of personal advice, and as a place to request a diagnosis\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. From the perspective of research staff, having study participants utilize SRH chatbots was seen as a way to reduce their workload and set boundaries outside of work hours\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Participants even perceived chatbots as \u0026ldquo;friends\u0026rdquo;, seeking out their advice when trusted individuals were not around to provide answers\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eAlthough AI-generated education was praised for being generated quickly\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, it required longer reading times than Google Search results\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e or clinician-generated responses\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. Many considered chatbots inferior to healthcare professionals or internet search results due to distrust and low perceived accuracy\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Using the Flesh-Kincaid score, a verified metric for readability of the written English language, one study showed that chatbot-based SRH information was written at the college graduate or professional level\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eImpact\u003c/h2\u003e \u003cp\u003eIn one study of clinicians, 22% perceived chatbot responses as effective and 24% perceived them as ineffective for SRH advice\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. From the participant perspective, the impact of chatbot-powered SRH education included participant-perceived increases in HIV and family planning knowledge\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e, positive impacts on intentions to get tested for STIs\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e, and subjective empowerment of young people to advocate for their sexual feelings and rights\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Chatbots were found to satisfy most users\u0026rsquo; queries, often dispelling misconceptions and providing accurate education\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e, and correctly counseling users to seek professional medical advice\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHowever, the impact of AI-based education was hindered by its occasional factual inaccuracy, excessive information, and poor organization\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. Not all AI chatbots are equally impactful\u0026mdash;one study found superior accuracy and completeness of answers provided by ChatGPT compared to Google Bard AI\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. Furthermore, AI-generated answers are often better at answering some SRH-related queries than others. One study found that information provided about PrEP and STI by a chatbot scored higher than clinician-generated responses, but its information about contraception was seen as inferior to human-created education.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003e This systematic review synthesizes the available literature on the use of artificial intelligence (AI) in sexual and reproductive health education, assessing its acceptability, feasibility, and impact, highlighting both the promises and the challenges associated with AI-driven sex education interventions.\u003c/p\u003e \u003cp\u003eThe acceptability of AI-based interventions in the healthcare space is based on user factors such as trust, literacy, and receptiveness; system factors such as workflow integration; and social factors such as ethics\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. Overall, AI-based sexual health education interventions appear to be widely accepted across diverse populations for providing immediate, accessible, and often judgment-free responses, making them particularly attractive to individuals seeking confidential sexual health information\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. The studies included in this review suggest that users appreciate the convenience, ease of use, and anonymity provided by AI chatbots, particularly for topics that may be stigmatized in traditional educational settings. However, concerns remain about the credibility, emotional disconnect, and readability that AI-based interventions can provide compared to trained sexual health educators or healthcare professionals\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe feasibility of AI-chatbots in public health interventions revolves around its ability to provide a non-judgemental space to discuss sensitive health information, and the potential for scalability\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. The feasibility of AI-driven sex education was supported by several studies. Chatbots demonstrated technical functionality, the ability to tailor responses to diverse user populations, ease of implementation fulfilling the role of quickly generating information, and the potential for reducing clinician workload. The ability of AI to deliver personalized educational content and respond to user queries in real-time enhances its potential as a scalable public health intervention. However, feasibility was limited in some studies by the variations in chatbot comprehension, accuracy, and reading times.\u003c/p\u003e \u003cp\u003eWith regard to impact, outcomes were mixed. Some studies included in this review indicate that AI chatbots can often provide factual sexual health information, and subjectively increased users\u0026rsquo; confidence discussing sexual feelings and needs, and created positive behavioral intentions. However, other studies revealed that some AI chatbots functioned better than other chatbots, and certain topics were better taught by artificial intelligence than other topics. Ultimately, the lack of rigorous randomized experimental designs limits the ability to draw definitive conclusions about AI\u0026rsquo;s efficacy, or even its superiority or inferiority compared to human-led sex education programs.\u003c/p\u003e \u003cp\u003eSeveral key recommendations emerge from this review. First, future research should prioritize the development and evaluation of AI-driven sex education interventions using high-quality study designs, including randomized controlled trials (RCTs) with appropriate control groups. Additionally, AI-driven interventions should be integrated into broader educational and healthcare systems to ensure that they complement, rather than replace, human-led education and counseling\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. Finally, further investigation is needed to assess the long-term efficacy of AI-based sex education on behavioral outcomes, including safer sex practices, STI prevention, and contraceptive use.\u003c/p\u003e \u003cp\u003e A major strength of this systematic review is its comprehensive evaluation of AI-driven sexual health education across multiple dimensions, including acceptability, feasibility, and impact. The inclusion of studies from diverse geographic regions and gray literature enhances the generalizability of the findings while reducing publication bias. Additionally, the use of a standardized quality assessment tool (MMAT) strengthens the validity of this review\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eSeveral limitations must be acknowledged. First, the lack of RCTs limit the ability to establish causal relationships between AI interventions and improved sexual health knowledge or behaviors, with all 21 studies lacking results that show clear efficacy. In addition, heterogeneity between study design and publication type intrinsically reduces the comparability of findings, preventing our ability to complete a meta-analysis. To standardize quality analysis, we utilized the Mixed-Methods Appraisal Tool to assess all studies on the basis of their individual study designs. Further, moderate inter-rater reliability in the screening round was resolved with additional reviewer training and discussion on a priori eligibility criteria. Additionally, large language models are trained by a limited scope and sometimes false or biased information, which has the potential to lead to knowledge gaps or misinformation in fields such as abortion and LGBTQ\u0026thinsp;+\u0026thinsp;health. Finally, the rapid evolution of AI technology means that some of the findings may become outdated as newer, more sophisticated AI models emerge.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eWhile AI has demonstrated promise in delivering accessible and user-friendly sexual health education, the current evidence-base remains limited. AI chatbots and virtual assistants have the potential to complement traditional sex education methods, particularly in addressing stigma and providing real-time information. However, significant gaps remain in assessing their research-based efficacy and real-world effectiveness relative to human-led approaches. Future research should focus on rigorously evaluating AI\u0026rsquo;s role in sexual health education to ensure that it is both a credible and effective tool for modern public health advancement.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eArtificial intelligence (AI)\u003c/p\u003e\n\u003cp\u003ePreferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA)\u003c/p\u003e\n\u003cp\u003eMixed Methods Appraisal Tool (MMAT)\u003c/p\u003e\n\u003cp\u003eRandomized Control Trial (RCT)\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cem\u003eEthics approval and consent to participate:\u0026nbsp;\u003c/em\u003eNot applicable\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eConsent for publication:\u0026nbsp;\u003c/em\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAvailability of data and materials:\u003c/em\u003e Template data collection forms and data extracted from the included studies are available by request\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCompeting interests:\u0026nbsp;\u003c/em\u003eThe authors report there are no competing interests to declare.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eFunding:\u0026nbsp;\u003c/em\u003eThis work was supported by the 2024 Lazarus Family Scholarship at George Washington University.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAuthors\u0026apos; Contributions:\u0026nbsp;\u003c/em\u003eStudy conceptualization was led by SB under the guidance of BCZ. SB and CB participated in data search, screening, data extraction, and quality assessment. SB led data synthesis and preparation of the manuscript, tables, and figures. All authors edited and approved the final manuscript for submission.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAcknowledgements:\u0026nbsp;\u003c/em\u003eThank you to Ms. Laura Abate and Mr. Tom Harrod, librarians at the George Washington University School of Medicine and Health Sciences, for their guidance on systematic review methods.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLameiras-Fern\u0026aacute;ndez M, Mart\u0026iacute;nez-Rom\u0026aacute;n R, Carrera-Fern\u0026aacute;ndez MV, Rodr\u0026iacute;guez-Castro Y. Sex education in the spotlight: what Is working? Systematic review. Int J Environ Res Public Health. 2021;18(5):2555. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/ijerph18052555\u003c/span\u003e\u003cspan address=\"10.3390/ijerph18052555\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 33806507; PMCID: PMC7967369.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBishop MD, Mallory AB, Gessner M, Frost DM, Russell ST. School-based sexuality education experiences across three generations of sexual minority people. J Sex Res. 2021 May-Jun;58(5):648\u0026ndash;658. doi: 10.1080/00224499.2020.1767024. Epub 2020 Jun 2. PMID: 32486928; PMCID: PMC7708407.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGray BJ, Jones AT, Couzens Z, Sagar T, Jones D. University students' behaviours towards accessing sexual health information and treatment. Int J STD AIDS. 2019;30(7):671\u0026ndash;679. doi: 10.1177/0956462419828866. Epub 2019 Apr 8. PMID: 30961468.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarbour AB, Frush JM, Gatta LA, McManigle WC, Keah NM, Bejarano-Pineda L, Guerrero EM. Artificial intelligence in health care: insights from an educational forum. J Med Educ Curric Dev. 2020;6:2382120519889348. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1177/2382120519889348\u003c/span\u003e\u003cspan address=\"10.1177/2382120519889348\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 32064356; PMCID: PMC6993147.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDave M, Patel N. Artificial intelligence in healthcare and education. Br Dent J. 2023;234, 761\u0026ndash;764. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41415-023-5845-2\u003c/span\u003e\u003cspan address=\"10.1038/s41415-023-5845-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYoung SD, Crowley JS, Vermund SH. Artificial intelligence and sexual health in the USA. Lancet Digit Health. 2021;3(8):e467-e468. doi: 10.1016/S2589-7500(21)00117-5. PMID: 34325852; PMCID: PMC10767714.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePage MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/bmj.n71\u003c/span\u003e\u003cspan address=\"10.1136/bmj.n71\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCovidence systematic review software, Veritas Health Innovation, Melbourne, Australia. Available at www.covidence.org\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHong QN, Pluye P, F\u0026agrave;bregues S, Bartlett G, Boardman F, Cargo M, et al. Mixed methods appraisal tool (MMAT) version 2018 user guide. McGill Department of Family Medicine. 2018 Aug 1.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCenters for Disease Control and Prevention (CDC). GRADE criteria for determining certainty of evidence [Internet]. Atlanta, GA: CDC; [cited 2025 Feb 27]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cdc.gov/acip-grade-handbook/hcp/chapter-7-grade-criteria-determining-certainty-of-evidence/index.html\u003c/span\u003e\u003cspan address=\"https://www.cdc.gov/acip-grade-handbook/hcp/chapter-7-grade-criteria-determining-certainty-of-evidence/index.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYam EA, Namukonda E, McClair T, Souidi S, Chelwa N, Muntalima N, et al. Developing and testing a chatbot to integrate HIV education into family planning clinic waiting areas in Lusaka, Zambia. Glob Health Sci Pract. 2022;10(5). doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.9745/GHSP-D-21-00721\u003c/span\u003e\u003cspan address=\"10.9745/GHSP-D-21-00721\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNadarzynski T, Knights N, Husbands D, Graham CA, Llewellyn CD, Buchanan T, et al. The impact of Chatbot-Assisted Self-Assessment (CASA) on intentions for sexual health screening in people from minoritised ethnic groups at risk of sexually transmitted infections. Sex Health. 2024;21. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1071/SH24058\u003c/span\u003e\u003cspan address=\"10.1071/SH24058\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang H, Gupta S, Singhal A, Muttreja P, Singh S, Sharma P, et al. An artificial intelligence chatbot for young people's sexual and reproductive health in India (SnehAI): instrumental case study. J Med Internet Res. 2022;24(1):e29969. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2196/29969\u003c/span\u003e\u003cspan address=\"10.2196/29969\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurns C, Bakaj A, Berishaj A, Hristidis V, Deak P, Equils O. Use of generative AI for improving health literacy in reproductive health: case study. JMIR Form Res. 2024;8:e59434. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2196/59434\u003c/span\u003e\u003cspan address=\"10.2196/59434\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHanai A, Ishikawa T, Kawauchi S, Iida Y, Kawakami E. Generative artificial intelligence and non-pharmacological bias: an experimental study on cancer patient sexual health communications. BMJ Health Care Inform. 2024;31(1). doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/bmjhci-2023-100924\u003c/span\u003e\u003cspan address=\"10.1136/bmjhci-2023-100924\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNjogu J, Jaworski G, Oduor C, Chea A, Malmqvist A, Rothschild CW. Assessing acceptability and effectiveness of a pleasure-oriented sexual and reproductive health chatbot in Kenya: an exploratory mixed-methods study. Sex Reprod Health Matters. 2023;31(4):2269008. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/26410397.2023.2269008\u003c/span\u003e\u003cspan address=\"10.1080/26410397.2023.2269008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNadarzynski T, Bayley J, Llewellyn C, Kidsley S, Graham CA. Acceptability of artificial intelligence (AI)-enabled chatbots, video consultations and live webchats as online platforms for sexual health advice. BMJ Sex Reprod Health. 2020;46(3):210\u0026ndash;217. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/bmjsrh-2018-200271\u003c/span\u003e\u003cspan address=\"10.1136/bmjsrh-2018-200271\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNadarzynski T, Lunt A, Knights N, Bayley J, Llewellyn C. \"But can chatbots understand sex?\" Attitudes towards artificial intelligence chatbots amongst sexual and reproductive health professionals: an exploratory mixed-methods study. Int J STD AIDS. 2023;34(11):809\u0026ndash;816. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1177/09564624231180777\u003c/span\u003e\u003cspan address=\"10.1177/09564624231180777\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNadarzynski T, Puentes V, Pawlak I, Mendes T, Montgomery I, Bayley J, et al. Barriers and facilitators to engagement with artificial intelligence (AI)-based chatbots for sexual and reproductive health advice: a qualitative analysis. Sex Health. 2021;18(5):385\u0026ndash;393. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1071/SH21123\u003c/span\u003e\u003cspan address=\"10.1071/SH21123\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGreen EP, Whitcomb A, Kahumbura C, Rosen JG, Goyal S, Achieng D, et al. \"What is the best method of family planning for me?\": a text mining analysis of messages between users and agents of a digital health service in Kenya. Gates Open Res. 2019;3:1475. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.12688/gatesopenres.12999.1\u003c/span\u003e\u003cspan address=\"10.12688/gatesopenres.12999.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNg\u0026ouml;njiri A, Memiah P, Kimathi R, Wagner FA, Ikahu A, Omanga E, et al. Utilizing user preferences in designing the AGILE (Accelerating Access to Gender-Based Violence Information and Services Leveraging on Technology Enhanced) chatbot. Int J Environ Res Public Health. 2023;20(21). doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/ijerph20217018\u003c/span\u003e\u003cspan address=\"10.3390/ijerph20217018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcMahon HV, McMahon BD. Automating untruths: ChatGPT, self-managed medication abortion, and the threat of misinformation in a post-Roe world. Front Digit Health. 2024;6:1287186. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fdgth.2024.1287186\u003c/span\u003e\u003cspan address=\"10.3389/fdgth.2024.1287186\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChandler RD, Warner S, Aidoo-Frimpong G, Wells J. \"What Did You Say, ChatGPT?\" The use of AI in Black women's HIV self-education: an inductive qualitative data analysis. J Assoc Nurses AIDS Care. 2024 May-Jun;35(3):294\u0026ndash;302. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/JNC.0000000000000468\u003c/span\u003e\u003cspan address=\"10.1097/JNC.0000000000000468\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMediboina A, Badam RK, Chodavarapu S. Assessing the accuracy of information on medication abortion: a comparative analysis of ChatGPT and Google Bard AI. Cureus. 2024;16(1):e51544. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7759/cureus.51544\u003c/span\u003e\u003cspan address=\"10.7759/cureus.51544\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eComulada WS, Rezai R, Sumstine S, Flores DD, Kerin T, Ocasio MA, et al. A necessary conversation to develop chatbots for HIV studies: qualitative findings from research staff, community advisory board members, and study participants. AIDS Care - Psychological and Socio-Medical Aspects of AIDS/HIV. 2024;36(4):463\u0026ndash;471. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/09540121.2023.2216926\u003c/span\u003e\u003cspan address=\"10.1080/09540121.2023.2216926\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiew TW, Tan S-M, Yoo NE, Gan CL, Lee YY. Let's talk about Sex!: AI and relational factors in the adoption of a chatbot conveying sexual and reproductive health information. Comput Hum Behav Rep. 2023;11:100323. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.chbr.2023.100323\u003c/span\u003e\u003cspan address=\"10.1016/j.chbr.2023.100323\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBonnevie E, Lloyd TD, Rosenberg SD, Williams K, Goldbarg J, Smyser J. Layla's Got You: Developing a tailored contraception chatbot for Black and Hispanic young women. Health Educ J. 2021;80(4):413\u0026ndash;424. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1177/0017896920981122\u003c/span\u003e\u003cspan address=\"10.1177/0017896920981122\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRahman R, Rahman MR, Tripto NI. Adolescentbot: Understanding opportunities for chatbots in combating adolescent sexual and reproductive health problems in Bangladesh. In: Conference on Human Factors in Computing Systems - Proceedings; 2021. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1145/3411764.3445694\u003c/span\u003e\u003cspan address=\"10.1145/3411764.3445694\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHussain SA, Ogundimu F, Bhattarai S. Mobile phone-based chatbot for family planning and contraceptive information. Lecture Notes in Computer Science; 2019. p. 342\u0026ndash;352. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-3-030-22219-2_26\u003c/span\u003e\u003cspan address=\"10.1007/978-3-030-22219-2_26\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaylor R, Messham R, Okafor C, Pink E, Portas L, Gees H. Can sexual health clinicians be replaced by robots? Utility of artificial intelligence platforms to give sexual health advice. Sex Transm Infect. 2023;99(Suppl 1):A8-A8. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/sextrans-BASHH-2023.16\u003c/span\u003e\u003cspan address=\"10.1136/sextrans-BASHH-2023.16\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSharevski F, Loop JV, Jachim P, Devine A, Pieroni E. Talking abortion (mis)information with ChatGPT on TikTok. In: Proceedings \u0026ndash;\u0026thinsp;8th IEEE European Symposium on Security and Privacy Workshops, Euro S and PW 2023; 2023; 594\u0026ndash;608. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/EuroSPW59978.2023.00071\u003c/span\u003e\u003cspan address=\"10.1109/EuroSPW59978.2023.00071\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMehta N, Gupta S, Kularathne Y. The role and impact of artificial intelligence in addressing sexually transmitted infections, nonvenereal genital diseases, sexual health, and wellness. Indian Dermatol Online J. 2023;14(6):793\u0026ndash;798. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.4103/idoj.idoj_426_23\u003c/span\u003e\u003cspan address=\"10.4103/idoj.idoj_426_23\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 38099049; PMCID: PMC10718125.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKerasidou A. Artificial intelligence and the ongoing need for empathy, compassion and trust in healthcare. Bull World Health Organ. 2020;98(4):245\u0026ndash;250. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2471/BLT.19.237198\u003c/span\u003e\u003cspan address=\"10.2471/BLT.19.237198\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2020 Jan 27. PMID: 32284647; PMCID: PMC7133472.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhai C, Wibowo S, Li LD. Evaluating the AI dialogue system's intercultural, humorous, and empathetic dimensions in English language learning: A case study. Comput Educ Artif Intell. 2024;7:100262.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaleki Varnosfaderani S, Forouzanfar M. The role of AI in hospitals and clinics: transforming healthcare in the 21st Century. Bioengineering (Basel). 2024;11(4):337. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/bioengineering11040337\u003c/span\u003e\u003cspan address=\"10.3390/bioengineering11040337\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 38671759; PMCID: PMC11047988.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHua D, Petrina N, Young N, Cho JG, Poon SK. Understanding the factors influencing acceptability of AI in medical imaging domains among healthcare professionals: A scoping review. Artif Intell Med. 2024;147:102698. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.artmed.2023.102698\u003c/span\u003e\u003cspan address=\"10.1016/j.artmed.2023.102698\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2023 Nov 9. PMID: 38184343.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAggarwal A, Tam CC, Wu D, Li X, Qiao S. Artificial Intelligence-Based Chatbots for Promoting Health Behavioral Changes: Systematic Review. J Med Internet Res. 2023;25:e40789. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2196/40789\u003c/span\u003e\u003cspan address=\"10.2196/40789\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 36826990; PMCID: PMC10007007.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTable 1 is available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"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":"artificial intelligence, sex education, sex ed, sexual health education, reproductive health, reproductive health education, chatbot, large language model, generative AI, AI","lastPublishedDoi":"10.21203/rs.3.rs-6289967/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6289967/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground:\u003c/strong\u003e Artificial intelligence (AI) is evolving and expanding at an unprecedented rate across healthcare and education. AI for sexual health education has the potential to reduce sexual health stigma, provide convenience for many populations of all genders, sexualities, and ages who were previously receiving insufficient or outdated information, and reduce the resources needed to provide this essential education. The aim of this systematic review is to assess the acceptability, feasibility, and impact of generative AI in sexual and reproductive health education.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods\u003c/strong\u003e: We searched PubMed, Web of Science, and Scopus in August 2024 combining artificial intelligence and sexual education search terms. We included experimental and observational studies of any analysis technique published between 01/01/2014-8/16/2024. Data was managed in Covidence. Screening and extraction utilized two non-expert reviewers. Quality assessment utilized the Mixed Methods Appraisal Tool and reporting adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003e Initial database search yielded 4,044 records, 21 full-text articles meeting inclusion criteria. All were observational studies. Data included 146,990 individual participants (mean=7000, median=100) from eight countries. Five (23.8%) compared an AI chatbot to another format of sex education. Eighteen studies assessed acceptability, 12 studies assessed feasibility, and 13 studies assessed impact. Users of AI primarily seek factual information, find the chatbot's responses easy to understand, and appreciate the immediate responses compared to human responses. AI helps users exercise sexual rights, discuss sexual feelings/needs, and learn information about HIV and family planning. However, chatbot responses differ in tone and empathy than human responses and require long reading times. While chatbots are generally viewed as clinically safe and hold potential for providing accessible sexual health information, users show skepticism about their credibility for sensitive topics compared to human interactions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions:\u003c/strong\u003e Usage of AI is surpassing high-quality evidence about its acceptability, feasibility, and impact. While initial studies show promise of AI chatbots for presenting sexual health information, high-quality, randomized studies with human participants and comparator groups are needed before AI can be trusted to successfully deliver such education.\u003c/p\u003e","manuscriptTitle":"Can AI Teach Sex Ed? A Systematic Review of the Use of Artificial Intelligence in Sexual and Reproductive Health Education","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-03-26 09:47:17","doi":"10.21203/rs.3.rs-6289967/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":"fae88c10-494e-44ff-9aa6-0a4e68cd1973","owner":[],"postedDate":"March 26th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-06-18T08:53:59+00:00","versionOfRecord":[],"versionCreatedAt":"2025-03-26 09:47:17","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6289967","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6289967","identity":"rs-6289967","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}