{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T19:08:13Z","timestamp":1782846493152,"version":"3.54.5"},"reference-count":31,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JMIR AI"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec sec-type=\"background\">\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Images created with generative artificial intelligence (AI) tools are increasingly used for health communication due to their ease of use, speed, accessibility, and low cost. However, AI-generated images may bring practical and ethical risks to health practitioners and the public, including through the perpetuation of stigma against vulnerable and historically marginalized groups.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"objective\">\n                    <jats:title>Objective<\/jats:title>\n                    <jats:p>To understand the potential value of AI-generated images for health care and public health communication, we sought to analyze images of substance use disorder and recovery generated with ChatGPT. Specifically, we sought to investigate: (1) the default visual outputs produced in response to a range of prompts about substance use disorder and recovery, and (2) the extent to which prompt modification and guideline-informed prompting could mitigate potentially stigmatizing imagery.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"methods\">\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We performed a mixed-methods case study examining depictions of substance use and recovery in images generated by ChatGPT 4.o. We generated images (n=84) using (1) prompts with colloquial and stigmatizing language, (2) prompts that follow best practices for person-first language, (3) image prompts written by ChatGPT, and (4) a custom GPT informed by guidelines for images of SUD. We then used a mixed-methods approach to analyze images for demographics and stigmatizing elements.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"results\">\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Images produced in the default ChatGPT model featured primarily White men (81%, n=34). Further, images tended to be stigmatizing, featuring injection drug use, dark colors, and symbolic elements such as chains. These trends persisted even when person-first language prompts were used. Images produced by the guideline-informed custom GPT were markedly less stigmatizing; however, they featured almost only Black women (74%, n=31).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec sec-type=\"conclusions\">\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>Our findings confirm prior research about stigma and biases in AI-generated images and extend this literature to substance use. However, our findings also suggest that (1) images can be improved when clear guidelines are provided and (2) even with guidelines, iteration is needed to create an image that fully concords with best practices.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.2196\/81977","type":"journal-article","created":{"date-parts":[[2026,2,19]],"date-time":"2026-02-19T20:42:45Z","timestamp":1771533765000},"page":"e81977-e81977","source":"Crossref","is-referenced-by-count":0,"title":["AI-Generated Images of Substance Use and Recovery: Mixed Methods Case Study"],"prefix":"10.2196","volume":"5","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-3074-2294","authenticated-orcid":false,"given":"Kathryn","family":"Heley","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-1613-4400","authenticated-orcid":false,"given":"Jeffrey K","family":"Hom","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3272-9317","authenticated-orcid":false,"given":"Linnea","family":"Laestadius","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2026,2,19]]},"reference":[{"issue":"6","key":"R1","doi-asserted-by":"publisher","first-page":"776","DOI":"10.1377\/hlthaff.2024.00050","article-title":"Transforming public health practice with generative artificial intelligence","volume":"43","author":"Bharel","journal-title":"Health Aff (Millwood)"},{"issue":"10","key":"R2","doi-asserted-by":"publisher","first-page":"1693","DOI":"10.1080\/21670811.2024.2331769","article-title":"Generative visual AI in news organizations: challenges, opportunities, perceptions, and policies","volume":"13","author":"Thomson","journal-title":"Digital Journalism"},{"key":"R3","doi-asserted-by":"publisher","DOI":"10.5888\/pcd14.170017","article-title":"Public health 3.0: a call to action for public health to meet the challenges of the 21st century","volume":"14","author":"DeSalvo","journal-title":"Prev Chronic Dis"},{"issue":"2","key":"R4","doi-asserted-by":"publisher","first-page":"145","DOI":"10.34197\/ats-scholar.2022-0141PS","article-title":"A picture worth a thousand words, created with one sentence: using artificial intelligence-created art to enhance medical education","volume":"4","author":"Huston","journal-title":"ATS Sch"},{"issue":"2","key":"R5","doi-asserted-by":"publisher","first-page":"33","DOI":"10.1192\/bjp.2023.136","article-title":"Artificial intelligence and increasing misinformation","volume":"224","author":"Monteith","journal-title":"Br J Psychiatry"},{"issue":"1","key":"R6","doi-asserted-by":"publisher","DOI":"10.1136\/bmjment-2024-301298","article-title":"AI depictions of psychiatric diagnoses: a preliminary study of generative image outputs in Midjourney V.6 and DALL-E 3","volume":"27","author":"Flathers","journal-title":"BMJ Ment Health"},{"issue":"8","key":"R7","doi-asserted-by":"publisher","first-page":"2563","DOI":"10.1007\/s00406-025-01998-x","article-title":"Biased AI generated images of mental illness: does AI adopt our stigma?","volume":"275","author":"Papazova","journal-title":"Eur Arch Psychiatry Clin Neurosci"},{"key":"R8","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.findings-naacl.266","article-title":"Decoding fatphobia: examining anti-fat and pro-thin bias in AI-generated images","author":"Warren","journal-title":"Findings of the Association for Computational Linguistics"},{"issue":"1","key":"R9","doi-asserted-by":"publisher","DOI":"10.2196\/56891","article-title":"Generative AI in medicine: pioneering progress or perpetuating historical inaccuracies? Cross-Sectional Study Evaluating Implicit Bias","volume":"4","author":"Sutera","journal-title":"JMIR AI"},{"issue":"11","key":"R10","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pmed.1002969","article-title":"Stigma as a fundamental hindrance to the United States opioid overdose crisis response","volume":"16","author":"Tsai","journal-title":"PLoS Med"},{"issue":"10","key":"R11","doi-asserted-by":"publisher","first-page":"2005","DOI":"10.1111\/add.70122","article-title":"Healthcare provider stigma toward patients with substance use disorders","volume":"120","author":"Parish","journal-title":"Addiction"},{"issue":"4","key":"R12","doi-asserted-by":"publisher","first-page":"919","DOI":"10.1080\/08897077.2021.1900988","article-title":"Pharmacists\u2019 stigma toward patients engaged in opioid misuse: when \u201csocial distance\u201d does not mean disease prevention","volume":"42","author":"Werremeyer","journal-title":"Subst Abuse"},{"issue":"1","key":"R13","doi-asserted-by":"publisher","DOI":"10.1186\/s12875-019-1047-z","article-title":"Attitudes of primary care physicians toward prescribing buprenorphine: a narrative review","volume":"20","author":"Louie","journal-title":"BMC Fam Pract"},{"issue":"5","key":"R14","doi-asserted-by":"publisher","DOI":"10.1258\/shorts.2012.011158","article-title":"The constructive use of images in medical teaching: a literature review","volume":"3","author":"Norris","journal-title":"JRSM Short Rep"},{"key":"R15","unstructured":"Alcohol Tobacco and other Drugs Council Tasmania. ATDC Image Guidelines. URL: https:\/\/atdc.org.au\/wp-content\/uploads\/2024\/01\/ATDC-Image-Guidelines-ATOD.pdf [Accessed 05-07-2025]"},{"key":"R16","unstructured":"Images matter: a guide to reducing stigma in visual media. Denver Health CAM Academy. 2025. URL: https:\/\/www.denverhealth.org\/sites\/default\/files\/2025-06\/CAM2504-56-Images-Matter-Flyer_C.pdf [Accessed 05-07-2025]"},{"key":"R17","unstructured":"Empathy Lens Stigma in images: How to avoid stereotypes and humanize people who use drugs. URL: https:\/\/empathylens.org\/wp-content\/uploads\/stigma-images-brochure.pdf [Accessed 25-07-2025]"},{"issue":"1","key":"R18","doi-asserted-by":"publisher","DOI":"10.1186\/s40352-023-00229-6","article-title":"Stigmatizing imagery for substance use disorders: a qualitative exploration","volume":"11","author":"Hulsey","journal-title":"Health Justice"},{"key":"R19","unstructured":"Creating a repository of non-stigmatizing visual images on substance use. The Ontario Network of People who Use Drugs, Public Health Ontario, Healthcare Human Factors. Mar 2024. URL: https:\/\/www.catie.ca\/sites\/default\/files\/2025-02\/report-repository-non-stigmatizing-visual-images-substance-use-EN-2024.pdf [Accessed 05-07-2025]"},{"key":"R20","doi-asserted-by":"publisher","DOI":"10.1016\/j.ypmed.2025.108293","article-title":"The impact of news photographs on drug-related stigma: a randomized message testing experiment in a National Sample of US Adults","volume":"195","author":"Hom","journal-title":"Prev Med"},{"key":"R21","doi-asserted-by":"publisher","DOI":"10.1016\/j.drugalcdep.2018.05.005","article-title":"Substance use, recovery, and linguistics: the impact of word choice on explicit and implicit bias","volume":"189","author":"Ashford","journal-title":"Drug Alcohol Depend"},{"issue":"3","key":"R22","doi-asserted-by":"publisher","first-page":"202","DOI":"10.1016\/j.drugpo.2009.10.010","article-title":"Does it matter how we refer to individuals with substance-related conditions? A randomized study of two commonly used terms","volume":"21","author":"Kelly","journal-title":"Int J Drug Policy"},{"key":"R23","unstructured":"Creswell JW Clark VLP . Designing and Conducting Mixed Methods Research. 3rd ed. SAGE Publications, Inc; 2018."},{"key":"R24","unstructured":"Yin RK . Case Study Research and Applications: Design and Methods. 6th ed. SAGE Publications, Inc; 2018."},{"issue":"2","key":"R25","doi-asserted-by":"publisher","first-page":"243","DOI":"10.1177\/1094428115610808","article-title":"Research design for mixed methods: a triangulation-based framework and roadmap","volume":"20","author":"Turner","journal-title":"Organizational Research Methods"},{"issue":"1","key":"R26","doi-asserted-by":"publisher","first-page":"127","DOI":"10.1080\/08897077.2020.1748167","article-title":"changes and the use of addiction-related stigmatizing terms in news media","volume":"43","author":"Bessette","journal-title":"Subst Abus"},{"issue":"1","key":"R27","doi-asserted-by":"publisher","first-page":"159","DOI":"10.2307\/2529310","article-title":"The measurement of observer agreement for categorical data","volume":"33","author":"Landis","journal-title":"Biometrics"},{"key":"R28","doi-asserted-by":"publisher","DOI":"10.1016\/j.drugpo.2025.105104","article-title":"When words and images diverge: challenges of AI-generated drug use representations","volume":"147","author":"Hom","journal-title":"Int J Drug Policy"},{"issue":"1","key":"R29","doi-asserted-by":"publisher","first-page":"27","DOI":"10.1080\/1556035X.2016.1272071","article-title":"Recovery identity and wellbeing: is it better to be \u2018recovered\u2019 or \u2018in recovery\u2019?","volume":"12","author":"Best","journal-title":"J Groups Addict Recover"},{"key":"R30","unstructured":"Ananya . AI image generators often give racist and sexist results: can they be fixed? Nature New Biol. Mar 28, 2024. URL: https:\/\/www.nature.com\/articles\/d41586-024-00674-9 [Accessed 09-02-2026]"},{"issue":"4","key":"R31","doi-asserted-by":"publisher","first-page":"514","DOI":"10.1177\/10755470221113833","article-title":"Missing the bigger picture: the need for more research on visual health misinformation","volume":"44","author":"Heley","journal-title":"Sci Commun"}],"container-title":["JMIR AI"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/ai.jmir.org\/2026\/1\/e81977","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T18:15:56Z","timestamp":1782843356000},"score":1,"resource":{"primary":{"URL":"https:\/\/ai.jmir.org\/2026\/1\/e81977"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,2,19]]},"references-count":31,"URL":"https:\/\/doi.org\/10.2196\/81977","relation":{},"ISSN":["2817-1705"],"issn-type":[{"value":"2817-1705","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,2,19]]},"article-number":"v5i6e81977"}}