{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,7]],"date-time":"2026-10-07T16:25:28Z","timestamp":1791390328459,"version":"4.3.3"},"reference-count":41,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["J Med Internet Res"],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Youth are increasingly engaging with generative AI (GenAI) chatbots; yet, little is known about the nature of their interactions.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Objective<\/jats:title>\n                    <jats:p>This study used naturalistic behavioral observations of keystrokes typed into GenAI mobile apps to characterize the topics youth discuss with chatbots, describing themes by age and app, word count by theme (a proxy for quantity of engagement), and co-occurring themes.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>Participants were 3363 US youth aged 1-17 years using the Aura parental monitoring app, a commercially available service that guardians install on children\u2019s mobile devices for digital safety. Participants included only those who opted into the overlay Aura keyboard (to collect keystroke data) and who engaged in some degree of GenAI use. Participants\u2019 keystrokes for distinct GenAI apps by day (ie, user-app-days) were coded using a large language model with human review.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Usage patterns were highly variable, with most users accessing only 1 to 2 GenAI apps (range 1-15) across a median of 4 days (IQR 1-11, range 1-247; 0.4%-90% of days during the study period), sending a median of 6 (IQR 2-27, range 2-2865) messages per app-day. Results identified 11 mutually inclusive themes, including functional tool use; social practice; role-play, including general, romantic, and sexual; emotional support; friend-like interactions; and violence. Most user-app-days involved using AI as a tool (26,343\/42,355, 62%), but notable proportions included violence (6460\/42,355, 15%), sexual role-play (6393\/42,355, 15%), romantic role-play (3538\/42,355, 8%), general role-play (3707\/42,355, 9%), and friend-like interaction (2624\/42,355, 6%). Theme prevalence varied by age: tool use was more common among older adolescents, whereas violence and sexual or romantic role-play were relatively more common among younger users. Themes also varied by app, with general-purpose apps most often used as tools and companionship or role-play apps showing higher usage of sexual, romantic, general role-play, and violent content; however, ChatGPT appeared across all themes due to high overall use. Word counts were highest for user-app-days involving violence, followed by sexual role-play, romantic role-play, and emotional support, suggesting greater engagement when using GenAI in these ways. Co-occurrence analyses showed that tool use often appeared alone, whereas violence, sexual role-play, and romantic role-play frequently co-occurred within a user-app-day, including instances suggestive of violent sexual role-play.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>Findings suggest that youth engagement with GenAI varies substantially across users. Tool use is most common, requiring more nuanced investigation into the precise ways youth use GenAI to potentially support or thwart their development. GenAI engagement also may involve developmentally salient and potentially risky themes of sexuality, romance, and violence, especially among younger users. More research is needed that pairs GenAI engagement with psychological measures of well-being to examine the impacts of AI interactions and promote safety-focused design features.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.2196\/95819","type":"journal-article","created":{"date-parts":[[2026,6,25]],"date-time":"2026-06-25T14:15:05Z","timestamp":1782396905000},"page":"e95819","source":"Crossref","is-referenced-by-count":2,"title":["How US Youth Use AI Chatbots: Conversation Patterns From Naturalistic Keystroke Observations"],"prefix":"10.2196","volume":"28","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-1308-5798","authenticated-orcid":false,"given":"Anne J","family":"Maheux","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6127-592X","authenticated-orcid":false,"given":"Debra","family":"Boeldt","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1495-6399","authenticated-orcid":false,"given":"Samir","family":"Akre-Bhide","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0710-2512","authenticated-orcid":false,"given":"Jessica","family":"Flannery","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0002-6847-2074","authenticated-orcid":false,"given":"Allison","family":"Paige","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-6982-7446","authenticated-orcid":false,"given":"Giavanna","family":"Villella","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0959-6632","authenticated-orcid":false,"given":"Kaitlyn","family":"Burnell","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2351-6339","authenticated-orcid":false,"given":"Eva H","family":"Telzer","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6847-6935","authenticated-orcid":false,"given":"Scott H","family":"Kollins","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2026,7,27]]},"reference":[{"key":"ref1","unstructured":"Teen and young adult perspectives on generative AI: patterns of use, excitements, and concernsHopelab2026-07-13https:\/\/hopelab.org\/stories\/teen-and-young-adult-perspectives-on-generative-ai"},{"key":"ref2","unstructured":"MaddenMCalvinAHasseALenhartAThe Dawn of the AI Era: teens, parents, and the adoption of generative AI at home and schoolCommon Sense Media2026-07-13https:\/\/www.commonsensemedia.org\/sites\/default\/files\/research\/report\/2024-the-dawn-of-the-ai-era_final-release-for-web.pdf"},{"key":"ref3","unstructured":"SidotiOMcClainC34% of US adults have used ChatGPT, about double the share in 2023Pew Research Center2025-09-01https:\/\/www.pewresearch.org\/short-reads\/2025\/06\/25\/34-of-us-adults-have-used-chatgpt-about-double-the-share-in-2023\/"},{"key":"ref4","unstructured":"Artificial intelligence and adolescent well-being: an APA health advisoryAmerican Psychological Association20252026-07-13https:\/\/www.apa.org\/topics\/artificial-intelligence-machine-learning\/health-advisory-ai-adolescent-well-being.pdf"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/3757548"},{"key":"ref6","unstructured":"SingletonTGerkenTMcMahonLHow a chatbot encouraged a man who wanted to kill the QueenBBC News2025-10-12https:\/\/www.bbc.com\/news\/technology-67012224"},{"key":"ref7","unstructured":"HillKA teen was suicidal. ChatGPT was the friend he confided inThe New York Times2025-10-12https:\/\/www.nytimes.com\/2025\/08\/26\/technology\/chatgpt-openai-suicide.html"},{"key":"ref8","unstructured":"BickABlandinADemingDThe rapid adoption of generative AINational Bureau of Economic Research Working Paper 329662026-07-13http:\/\/www.nber.org\/papers\/w32966"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1007\/s13347-025-00863-y"},{"key":"ref10","unstructured":"FaverioMSidotiOTeens, social media and AI chatbots 2025Pew Research Center2026-07-13https:\/\/www.pewresearch.org\/wp-content\/uploads\/sites\/20\/2025\/12\/PI_2025.12.09_Teens-Social-Media-AI_REPORT.pdf"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1001\/jamanetworkopen.2025.56631"},{"key":"ref12","unstructured":"RobbMBMannSTalk, trust and trade-offs: how and why teens use AI companionsCommon Sense Media2026-07-13https:\/\/www.commonsensemedia.org\/sites\/default\/files\/research\/report\/talk-trust-and-trade-offs_2025_web.pdf"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1038\/s41562-021-01117-5"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1177\/21677026221125180"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1111\/jora.12541"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1111\/jcpp.13931"},{"key":"ref17","unstructured":"Online wellbeing by Aura: data on your kid\u2019s online habitsAura2026-07-13https:\/\/www.aura.com\/online-wellbeing"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1811.12808"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1177\/1609406920969268"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.21105\/joss.03021"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1080\/10494820.2024.2350655"},{"key":"ref22","unstructured":"Ten fresh insights on generative AI from our teen advisorsCenter for Digital Thriving2026-05-01https:\/\/digitalthriving.gse.harvard.edu\/wp-content\/uploads\/2025\/05\/CDT_Youth-Advisory-Insights_Gen-AI-Memo.pdf"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.31234\/osf.io\/pgyfc_v1"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1145\/3706598.3713429"},{"key":"ref25","author":"Bandura, A","year":"2009","journal-title":"Media Effects"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1016\/j.jpeds.2008.11.033"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1016\/S0140-6736(05)17952-5"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1016\/j.jecp.2020.105005"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1016\/j.chb.2012.11.001"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1080\/00224499.2016.1143441"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1007\/s11930-024-00397-y"},{"key":"ref32","unstructured":"JamaliLMcMahonLChatGPT will soon allow erotica for verified adults, OpenAI boss saysBBC News2025-12-08https:\/\/www.bbc.com\/news\/articles\/cpd2qv58yl5o"},{"key":"ref33","unstructured":"FriedIOpenAI delays ChatGPT \u201cadult mode\u201d and eroticaAxios20262026-04-29https:\/\/www.axios.com\/2026\/03\/06\/openai-delays-chatgpt-adult-mode"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.2139\/ssrn.4601555"},{"key":"ref35","unstructured":"CongerKFreedmanDThompsonSMusk's chatbot flooded X with millions of sexualized images in days, new estimates showThe New York Times20262026-02-06https:\/\/www.nytimes.com\/2026\/01\/22\/technology\/grok-x-ai-elon-musk-deepfakes.html"},{"key":"ref36","unstructured":"RaedlerJSwaroopSPanWAI companions are not the solution to loneliness: design choices and their drawbacks2025HAIC Workshop, International Conference on Learning Representations (ICLR 2025)April 27, 2025Singapore"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-026-69010-1"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1002\/jad.70164"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1111\/jcpp.14085"},{"key":"ref40","unstructured":"AndersonMParents, teens and digital monitoringPew Research Center2026-07-13https:\/\/www.pewresearch.org\/internet\/2016\/01\/07\/parents-teens-and-digital-monitoring\/"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.2196\/69358"}],"container-title":["Journal of Medical Internet Research"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.jmir.org\/2026\/1\/e95819","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,27]],"date-time":"2026-07-27T19:20:07Z","timestamp":1785180007000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.jmir.org\/2026\/1\/e95819"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,27]]},"references-count":41,"URL":"https:\/\/doi.org\/10.2196\/95819","relation":{"has-preprint":[{"id-type":"doi","id":"10.2196\/preprints.95819","asserted-by":"object"}]},"ISSN":["1438-8871"],"issn-type":[{"value":"1438-8871","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,27]]}}}