Teens Who Self-Injure Are Already Using AI to Cope.

interviews · Frontiers in Digital Health · la publicación, 26 ago 2026 · gratis

Lo que estos adolescentes dicen que la IA sí puede hacer por ellos, y dónde falla

Veintiún jóvenes estadounidenses que se autolesionaban contaron cómo usan el celular para sobrellevarlo. Sus respuestas sirven para conversar en casa, no como prueba de que alguna herramienta funcione.

Versión breve · la versión detallada sigue, unos 9 min

Pregunte a weeklyAI

Pregúnteme por este estudio: a quiénes se estudió, qué encontró y qué no dice.

Las conversaciones se guardan mientras exista weeklyAI, para mejorar la publicación. Se responde en el idioma en que usted escribe.

El estudio, de un vistazo
Quiénes: Adolescentes de 14 a 18 años que se habían autolesionado
Cuántos: 21 adolescentes
Dónde: Estados Unidos
Cuándo: Entrevistas entre mayo de 2024 y abril de 2025
Tipo de estudio: entrevistas
Quién lo hizo: Northwestern University Feinberg School of Medicine, University of California Irvine, Mental Health America
El límite que importa: Son 21 jóvenes que llegaron por su cuenta a una organización de salud mental; no mide si una herramienta funciona.

Si comparte algo sobre autolesión, la reacción importa: estos jóvenes pidieron curiosidad y calma, no alarma.
Lectura de weeklyAI

Ver · 0:52

Veintiún adolescentes estadounidenses de entre 14 y 18 años, que se habían autolesionado al menos dos días en el mes anterior, aceptaron ser entrevistados sobre cómo manejaban su malestar, a quién pedían ayuda y qué hacían con el celular.

Los entrevistó una trabajadora social clínica con licencia, entre mayo de 2024 y abril de 2025. Las conversaciones se grabaron, se transcribieron y se analizaron por temas. Los jóvenes llegaron al estudio a través de un cuestionario en el sitio de una organización nacional de salud mental.

Contaron que autolesionarse era su manera de sobrellevar la presión escolar, los conflictos y las exigencias familiares y culturales. Sabían que les hacía daño, pero muchos dudaban de querer dejarlo, y solían ocultarlo por vergüenza y por miedo a que un terapeuta tuviera que avisar a sus padres.

Para sostenerse, usaban de todo: música en plataformas como Spotify, aplicaciones para anotar el ánimo y escribir un diario, comunidades en línea y sistemas de inteligencia artificial como ChatGPT.

Lo que valoraban era concreto: está disponible a cualquier hora, es privado y responde de inmediato. Algunos dijeron que era más fácil abrirse con un programa que con una persona, porque no sentían que estuvieran molestando a nadie y no temían ser juzgados.

También describieron lo que falla. La IA daba respuestas genéricas —"confía en un amigo", "escribe un diario"— justo en los momentos más intensos, cuando querían algo específico para su situación. A veces respondía con errores. Y cuando mencionaban la autolesión, los filtros de seguridad cortaban la conversación, aunque necesitaran hablarlo. Las comunidades en línea, que aliviaban en el corto plazo, podían volverse un espacio donde el comportamiento se normalizaba o se volvía competitivo. La música ayudaba a distraerse, pero ciertas canciones intensificaban lo que ya sentían.

Son solo 21 jóvenes, todos llegados por su propia voluntad a una organización de salud mental, así que probablemente estén más atentos a su bienestar y más abiertos a lo digital que la mayoría. El estudio recogió lo que dijeron en entrevistas, no midió si alguna aplicación o chatbot reduce la autolesión. Nada de esto demuestra que una herramienta funcione, ni que sea segura, ni que reemplace la atención profesional.

Además, las entrevistas terminaron en abril de 2025 y la IA cambia muy rápido: lo que estos jóvenes describieron puede no coincidir con lo que su hijo o hija encuentra hoy.

Lo útil es preguntar sin dramatizar. Qué usa, para qué, y qué le deja eso. Si comparte algo sobre autolesión, la reacción importa: estos jóvenes pidieron curiosidad y calma, no alarma. Una pregunta para empezar hoy mismo: ¿qué haces con el celular cuando estás mal, y eso te ayuda o te deja peor?

Qué significa para usted

Ninguna herramienta sustituye a un profesional, y este estudio no probó que alguna reduzca la autolesión. Usted puede observar qué usa su hijo cuando está mal y preguntarle si eso lo ayuda o lo deja peor, sin alarmarse. Si menciona autolesión, escuche con calma y curiosidad; para decisiones de salud, consulte a un profesional.

Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943

Quién pagó: El trabajo fue financiado por una subvención del National Institute of Mental Health (K01MH131898); el artículo no indica si el financiador tuvo alguna función, y no menciona equipos ni programas cedidos por terceros.

No tome esto como consejo médico profesional.

Versión detalladaLos pasajes copiados del artículo, las ilustraciones y cada fuente con cuánto leímos de ella · 1777 palabras · unos 9 minLeerla →Cerrar

Los adolescentes que se lastiman ya usan ChatGPT, música y comunidades en línea para sobrellevarlo; un estudio entrevistó a 21 de ellos y muestra qué ayuda y qué empeora

Lo que dijeron no mide resultados: describe cómo se apoyan hoy, y por qué les cuesta contarlo en casa.

Así podría verse · ilustración generada por weeklyAI.watch, no es una fotografía

Un estudio publicado en la revista Frontiers in Digital Health entrevistó a 21 adolescentes estadounidenses de entre 14 y 18 años que se habían lastimado en el último mes, para entender cómo viven esa conducta, a quién le piden ayuda y qué hacen con la tecnología cuando están mal¹. No es un ensayo clínico ni una encuesta nacional: es un puñado de conversaciones largas, analizadas por temas, y sus hallazgos valen como descripción, no como prueba de que ninguna aplicación funcione¹.

En esas entrevistas, los adolescentes describieron la autolesión como una forma de sobrellevar emociones y problemas sociales más amplios, no como un fin en sí mismo². Sabían que les hace daño y, aun así, muchos se mostraron ambivalentes sobre dejarla: querer parar y no sentirse capaces de parar pueden convivir en la misma persona².

El estigma y el miedo a que otros se enteraran limitaban la búsqueda de ayuda formal, y eso los empujó hacia lo digital para informarse y recibir apoyo³. Allí usaban plataformas de música, aplicaciones de diario y de registro del ánimo, comunidades en línea y sistemas de inteligencia artificial como ChatGPT para regular emociones, reflexionar, pedir consejo y hablar de temas delicados⁴. Valoran que esas herramientas estén disponibles de inmediato, que sean privadas y que no hagan esperar; también señalaron límites: respuestas de la IA imprecisas o genéricas, poca moderación en las comunidades en línea y efectos emocionales complejos de la música⁵.

El atractivo de un chatbot aparece con una franqueza que sorprende. Una participante dijo que hablaba con ChatGPT y lo encontraba bastante útil⁶. Otra lo describió como su "pequeña terapeuta en línea" y explicó que le resultaba más cómodo que confiar en alguien conocido⁷. Otra más dijo que era casi más fácil abrirse porque sabía que no era humano⁸, y una cuarta lo resumió así: uno puede desahogarse sin culpa con un bot porque no está molestando a nadie⁹.

El problema es que esas conversaciones se cortan justo cuando más se necesitan. Una participante contó que los modelos de IA orientados a terapia bloqueaban el tema de la autolesión y se negaban a seguir la conversación, aunque ella necesitaba hablarlo¹⁰. Otra pidió respuestas concretas para un momento de crisis intensa y recibió lo mismo de siempre: confía en un amigo, escribe un diario, esos recordatorios generales que ya conocía de sobra¹¹.

Las comunidades en línea también tienen dos caras. Una participante describió una competencia silenciosa: ver a alguien que se lastima más o con heridas más graves puede hacer sentir que uno "no lo hace bien", y esas amistades no están enfocadas en la recuperación, así que todo el grupo se queda sin recuperarse y la conducta se vuelve normal¹². Otra dijo que hablar del tema en esos espacios ayuda a corto plazo pero empeora las cosas a largo plazo, y lo comparó con poner una curita sobre una herida de bala¹³.

Para dimensionar el fenómeno: el artículo recoge un estudio en el que alrededor del 17 por ciento de los adolescentes se lastima sin intención suicida, y la conducta se asocia con otros problemas de salud mental y con mayor riesgo de suicidio¹⁴. La mayoría de ellos no recibe tratamiento formal, por el estigma, por el costo y por la falta de profesionales disponibles¹⁵. Ese hueco es exactamente el que empuja a las familias y a los propios jóvenes hacia soluciones digitales.

No es un fenómeno marginal ni exclusivo de quienes se lastiman. Según el resumen de un estudio sobre riesgos de los chatbots conversacionales, que solo pudimos leer en su resumen porque el trabajo completo está detrás de una suscripción, más del 60 por ciento de los jóvenes encuestados usaba un chatbot de este tipo, y un 11.4 por ciento lo hacía a diario o casi a diario¹⁶. En ese mismo resumen, el 47.1 por ciento reportó haberse topado con uno o más riesgos o daños concretos¹⁷.

Ese resumen también detalla cuáles: al 32.3 por ciento le pidieron información personal incómoda; el 23.1 por ciento se sintió manipulado o presionado; el 17.1 por ciento recibió información falsa; el 18.7 por ciento fue incentivado a actuar de forma poco ética o ilegal; el 15.2 por ciento fue empujado hacia conductas de riesgo; y el 14.7 y el 13.0 por ciento quedaron expuestos a mensajes sobre autolesión y sobre suicidio, respectivamente¹⁸. El resumen señala que los varones, los jóvenes heterosexuales, los blancos y los de 13 años reportaron más experiencias negativas¹⁹.

Otro trabajo, un sondeo nacional representativo con adolescentes y jóvenes de 12 a 21 años realizado en noviembre de 2025 —que también pudimos leer solo en su resumen—, encontró que el 19.2 por ciento había usado chatbots de IA para pedir consejo sobre salud mental²⁰. De ellos, el 42.8 por ciento lo hacía al menos una vez al mes y el 91.7 por ciento calificó el consejo como algo o muy útil²¹. Y un dato que debería hacer pensar a cualquier adulto: el 63.3 por ciento no le había contado a nadie que usaba un chatbot para esto²².

Ese mismo resumen muestra quién recurre más a estas herramientas: las mujeres más que los varones²³, los jóvenes de 18 a 21 años más que los de 12 a 14²⁴, y quienes habían hablado con un médico sobre su salud mental en los seis meses previos más que quienes no lo habían hecho²⁵. Sus autores concluyen que los chatbots ya forman parte del ecosistema de información sobre salud mental de los jóvenes, y que padres y clínicos deberían hablarlo de forma proactiva²⁶.

Un tercer trabajo, también leído solo en su resumen, analizó conversaciones reales de 3,363 jóvenes usuarios de una aplicación comercial de monitoreo parental²⁷. Encontró que el uso es muy variable y que la mayoría recurre a una o dos aplicaciones de IA²⁸, con temas que van desde el uso como herramienta hasta el juego de roles, el apoyo emocional y las interacciones de amistad²⁹. La mayor parte del uso era instrumental, pero aparecieron proporciones notables de contenido violento y de juego de roles sexual o romántico³⁰, y esos temas variaban según la edad y según la aplicación³¹³².

Así lo leemos nosotros: cuando alguien no puede contar lo que le pasa sin quedar expuesto, no deja de necesitar contarlo; cambia de canal. Lo que este estudio sugiere es que muchos adolescentes que se lastiman no están escondiendo el tema por falta de ganas de hablarlo, sino porque el precio de decirlo en voz alta —que avisen a sus padres, que se arme un escándalo, que los miren distinto— les parece más alto que el alivio que buscan. Si eso es cierto en su casa, usted probablemente no se enterará por una conversación directa, sino por señales indirectas: qué aplicaciones abre cuando está mal, con quién habla de noche, qué hace con el celular después de un día difícil. Sabremos que nos equivocamos si su hijo o hija cuenta sus impulsos y sus recaídas a un adulto de confianza sin miedo a las consecuencias, o si deja de usar estas herramientas en cuanto tiene a alguien cerca disponible.

Lo segundo que notamos es que abrirse con otra persona se vive como una deuda. Varios de estos adolescentes describieron hablar con una máquina como un alivio precisamente porque no molesta, no se cansa y no espera nada a cambio. Eso no convierte al chatbot en un terapeuta ni en un sustituto de la ayuda profesional, y el estudio no midió si mejora o empeora nada. Pero sí explica por qué a veces piden ayuda a un programa justo cuando están más solos. La consecuencia práctica para una familia es incómoda: si usted reacciona con alarma o con un interrogatorio cuando por fin le cuentan algo, le está enseñando que contarlo sale caro. Lo que puede hacer es lo contrario: decir por adelantado qué pasaría si algún día le cuenta algo difícil, para que sepa que hablarlo no le va a costar más de lo que ya le cuesta, y preguntarle por lo que hace con el teléfono cuando está mal en lugar de preguntarle directamente por la conducta.

Los propios participantes dijeron qué les daría confianza en una herramienta pensada para esto: inicio de sesión con huella o reconocimiento facial, verificación en dos pasos, pantallas de salida rápida o interfaces neutrales que no revelen para qué sirve la aplicación³³. También pidieron que no se etiquete el contenido con la palabra autolesión, porque esa palabra sola ya los expone si alguien mira la pantalla.

Conviene ser claro sobre los límites. Eran 21 adolescentes, reclutados a través del sitio web de una organización nacional de salud mental, y quienes responden ese tipo de convocatoria suelen estar más conscientes de su salud mental y más abiertos a lo digital que el promedio¹. La mayoría se identificó como blanca y como mujer, lo que limita cuánto se puede trasladar a otros grupos. Quedaron fuera los adolescentes con diagnósticos graves y con riesgo suicida alto, así que estos relatos no hablan de los casos de mayor riesgo. Y las entrevistas se hicieron entre mayo de 2024 y abril de 2025, un período en el que la IA cambió muy rápido: lo que estos jóvenes describieron puede ya no coincidir con lo que su hijo encuentra hoy.

Aquí está lo que usted puede hacer con todo esto. Pregúntele a su hijo o hija qué hace con el celular cuando está mal, sin interrogatorio y sin revisar nada a escondidas: qué aplicación abre, con quién habla, qué le sirve y qué no. Si le dice que habla con una inteligencia artificial, no lo trate como una traición ni como una ridiculez; pregúntele qué le da esa conversación que no encuentra en otro lado, y dígale con palabras simples que contarle algo difícil a usted no lo convierte en una carga. Antes de prohibir una aplicación, pregúntese si en su casa se puede disentir sin castigo, porque un adolescente que no puede pensar en voz alta lo que siente va a seguir pensándolo en algún lugar sin rastro. Y si algún día le cuenta que se está lastimando, escuche primero y no prometa lo que no puede cumplir sobre lo que tendrá que decirle a otros: la transparencia sobre qué se comparte y con quién es lo que estos jóvenes pidieron una y otra vez. ¿Sabe usted hoy qué abre su hijo en el teléfono cuando el día se le vuelve insoportable?

De dónde sale cada dato de contexto, y cuánto leímos de cada documento

Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "We conducted semi-structured interviews between May 2024 and April 2025 with 21 adolescents (14–18 years old) with lived experience of self-injury."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Adolescents described self-injury as a coping strategy embedded within broader emotional and social stressors. While they were aware of the harms, they commonly expressed ambivalence about stopping the behavior."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Stigma and concerns about disclosure often limited help-seeking and contributed to their use of digital technologies to access information and support."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Participants reported using a range of tools, including music streaming platforms, journaling and mood-tracking apps, online communities, and AI systems such as ChatGPT to regulate emotions, reflect on their experiences, seek advice, and discuss sensitive topics."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Although these technologies were valued for their accessibility, privacy, and immediacy, participants also highlighted limitations, including inaccurate or generic AI responses, limited moderation in online communities, and the complex emotional impacts of music."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "PID16 commented “I talk to ChatGPT. I'll admit it, I talk to ChatGPT sometimes. And honestly, that's pretty helpful for me”"
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "PID11 noted, “I message ChatGPT… it feels like it's my little online therapist, and that helps,” emphasizing that talking to an AI could feel more comfortable than confiding in someone they knew."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "PID16 explained, “It's almost easier to open up to, because you know it's not human.”"
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "PID12 shared this sentiment, “You can rant guilt free to a bot because you're not being annoying. But it's like, if it's another person, you can't talk to a person to this extent or expect them to be there to this extent.”"
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "PID11 described how safety restrictions often blocked important conversations they hoped to have: “I’ve used a few different AI therapy-specific models. But they’re like, a little bit more limited, or if I want to talk about maybe self-injury – it’ll just be like, you can't, we’re not talking about this – like, it doesn't want to continue that conversation, and sometimes I need to talk about it.”"
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "PID18 explained, “If I was having a particular struggle moment, a very intense moment, I wanted specific answers to my specific situation — [But] it was also like, ‘Confide in a friend or journal,’ and a lot of the other very general mental health reminders that I’ve been surrounded by in communities.”"
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "PID16 said “A lot of issues like self-harm, especially online, can even be competitive in online spaces. If you see someone who does it more than you, or has more severe injuries than you, you think, ‘wow, I’m not doing it right,’ or ‘I can't even hurt myself good enough’”. She went on to say “It's just like a loop – you can make friends on there, but they’re also not focusing on recovery either. So, it's all of you gathered, just not recovering, and it becomes normal.”"
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "PID21 reflected that visiting communities helped in the short term but had a negative effect in the long term: “Talking about it is keeping it from getting better, and encouraging other people to do it too. So I think things are just band-aids over a bullet wound.”"
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Nonsuicidal self-injury (NSSI)—hereafter referred to as “self-injury”—is reported by approximately 17% of adolescents and is associated with a range of co-occurring mental health challenges and elevated risk of suicide ( 1 , 2 )."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Yet, most adolescents that engage in self-injury do not receive formal treatment. Stigma surrounding mental health concerns, and self-injury in particular, as well as structural barriers such as treatment costs and limited access to providers, contribute to this gap and highlight a need for alternative forms of intervention."
Hinduja S, Patchin JW. (2026). Risks and Harms of Conversational Artificial Intelligence (CAI) Chatbot Use Among US Youth. Journal of Adolescence. 10.1002/jad.70164 — solo el resumen - el texto completo está tras una suscripción — el pasaje: "Over 60% of the sample reported using a CAI chatbot, with 11.4% doing so every day or nearly every day."
Hinduja S, Patchin JW. (2026). Risks and Harms of Conversational Artificial Intelligence (CAI) Chatbot Use Among US Youth. Journal of Adolescence. 10.1002/jad.70164 — solo el resumen - el texto completo está tras una suscripción — el pasaje: "CAI chatbot usage is common among US adolescents, with 47.1% reporting exposure to one or more specific risks and harms."
Hinduja S, Patchin JW. (2026). Risks and Harms of Conversational Artificial Intelligence (CAI) Chatbot Use Among US Youth. Journal of Adolescence. 10.1002/jad.70164 — solo el resumen - el texto completo está tras una suscripción — el pasaje: "Still, 32.3% were asked for uncomfortable personal information, 23.1% felt manipulated or pressured, 17.1% received false information, 18.7% were encouraged to act unethically or illegally, 15.2% were prompted to risky behaviors, and 14.7% and 13.0% were exposed to self-harm and suicidal messages, respectively."
Hinduja S, Patchin JW. (2026). Risks and Harms of Conversational Artificial Intelligence (CAI) Chatbot Use Among US Youth. Journal of Adolescence. 10.1002/jad.70164 — solo el resumen - el texto completo está tras una suscripción — el pasaje: "Male, heterosexual, white, and younger (13-year-old) youth reported higher rates of most negative experiences."
McBain RK, Cantor JH, Breslau J, Diliberti M, Zhang LA, Zhang F, et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "19.2% of adolescents and young adults (population-weighted n = 8 207 180) in 2025 reported having used AI chatbots for mental health advice."
McBain RK, Cantor JH, Breslau J, Diliberti M, Zhang LA, Zhang F, et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "Among those who sought advice from AI chatbots, 42.8% did so at least monthly, and 91.7% rated the advice as somewhat or very helpful."
McBain RK, Cantor JH, Breslau J, Diliberti M, Zhang LA, Zhang F, et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "Most adolescents reported they had not disclosed AI chatbot use for mental health advice to anyone (63.3%)."
McBain RK, Cantor JH, Breslau J, Diliberti M, Zhang LA, Zhang F, et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "Use of an AI chatbot for mental health advice was more common among females compared with males (adjusted odds ratio [aOR], 2.10; 95% CI, 1.36-3.23)"
McBain RK, Cantor JH, Breslau J, Diliberti M, Zhang LA, Zhang F, et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "respondents aged 18 to 21 years compared with respondents aged 12 to 14 years (aOR, 3.65; 95% CI, 1.98-6.74)"
McBain RK, Cantor JH, Breslau J, Diliberti M, Zhang LA, Zhang F, et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "those who had spoken with a physician about their mental health in the prior 6 months compared with those who had not (aOR, 1.89; 95% CI, 1.18-3.03)"
McBain RK, Cantor JH, Breslau J, Diliberti M, Zhang LA, Zhang F, et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "AI chatbots are already embedded in many youths' mental health information ecosystem, underscoring the need for parents and clinicians to proactively discuss chatbot use to promote safety, appro"
Maheux AJ, Boeldt D, Akre-Bhide S, Flannery J, Paige A, Villella G, et al. (2026). How US Youth Use AI Chatbots: Conversation Patterns From Naturalistic Keystroke Observations. Journal of Medical Internet Research. https://doi.org/10.2196/95819 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "Participants were 3363 US youth aged 1-17 years using the Aura parental monitoring app, a commercially available service that guardians install on children's mobile devices for digital safety."
Maheux AJ, Boeldt D, Akre-Bhide S, Flannery J, Paige A, Villella G, et al. (2026). How US Youth Use AI Chatbots: Conversation Patterns From Naturalistic Keystroke Observations. Journal of Medical Internet Research. https://doi.org/10.2196/95819 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "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."
Maheux AJ, Boeldt D, Akre-Bhide S, Flannery J, Paige A, Villella G, et al. (2026). How US Youth Use AI Chatbots: Conversation Patterns From Naturalistic Keystroke Observations. Journal of Medical Internet Research. https://doi.org/10.2196/95819 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "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."
Maheux AJ, Boeldt D, Akre-Bhide S, Flannery J, Paige A, Villella G, et al. (2026). How US Youth Use AI Chatbots: Conversation Patterns From Naturalistic Keystroke Observations. Journal of Medical Internet Research. https://doi.org/10.2196/95819 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "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%)."
Maheux AJ, Boeldt D, Akre-Bhide S, Flannery J, Paige A, Villella G, et al. (2026). How US Youth Use AI Chatbots: Conversation Patterns From Naturalistic Keystroke Observations. Journal of Medical Internet Research. https://doi.org/10.2196/95819 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "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."
Maheux AJ, Boeldt D, Akre-Bhide S, Flannery J, Paige A, Villella G, et al. (2026). How US Youth Use AI Chatbots: Conversation Patterns From Naturalistic Keystroke Observations. Journal of Medical Internet Research. https://doi.org/10.2196/95819 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "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."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Participants mentioned safeguards such as biometric login, two-factor authentication, quick-exit screens, or neutral “cover” interfaces may meaningfully increase uptake among adolescents who fear stigmatization."

No tome esto como consejo médico profesional.

Los hallazgos de otros estudios que aquí se mencionan los conocemos por este documento, que fue el que leímos; no abrimos cada uno de esos estudios.

interviews · Frontiers in Digital Health · the paper, 26 Aug 2026 · free

Teens Who Self-Injure Are Already Using AI to Cope.

Short version · the longer version follows, about 6 min

Ask weeklyAI

Ask me about this study: who was studied, what it found, and what it does not say.

Conversations are saved for as long as weeklyAI exists, to improve the publication. Answers come in the language you write in.

The study at a glance
Who: adolescents with lived experience of self-injury
How many: 21 adolescents
Where: United States
When: May 2024 to April 2025
Kind of study: interviews
Who did it: Northwestern University, University of California Irvine, Mental Health America
The limit that matters: Only 21 teens, recruited from a mental health website, so they may differ from most.

The study reports what these teenagers said, not what any tool did for them.
weeklyAI's reading

Watch · 0:49

The 21 teenagers had all hurt themselves in the past month. When researchers asked how they coped, the answers rarely started with a doctor or a therapist.

They talked about music playlists. Journaling apps. Mood trackers. Online forums. And, again and again, chatbots.

The small study, published in Frontiers in Digital Health, interviewed US adolescents ages 14 to 18 between May 2024 and April 2025. They were recruited through a national mental health nonprofit's website, and each had self-injured on at least two days in the previous month.

They described self-injury as a way of handling stress—school pressure, family expectations, cultural demands—not as a problem in itself. Many were unsure they wanted to stop. Stigma, and fear that a therapist would have to tell their parents, kept them from seeking formal help.

So they turned to their phones.

Some used apps built for self-injury, like Calm Harm, which offers a button for moments of urge. Others journaled, tracked moods, or listened to music—though some said certain songs made urges worse rather than better.

Online communities offered connection and coping ideas, but also competition over who hurt themselves more, and content that encouraged the behavior.

Then there was AI. Teens said they talked to ChatGPT about feelings they couldn't bring themselves to share with people. They valued that it was always available and wouldn't judge them.

But they also described generic advice—"confide in a friend or journal"—and safety filters that shut down conversations about self-injury entirely. Some got responses that were plainly wrong.

The study reports what these teenagers said, not what any tool did for them. Nothing was measured over time. No app or chatbot has been shown here to reduce self-injury.

The teens were also self-selected, recruited through a mental health website, so they may be more aware and more motivated than most.

And because the interviews ended in April 2025, what your child encounters today may already look different.

That makes one question worth asking at your kitchen table: *What do you use, and what does it actually do for you?*

What this means for you

What the teenagers described is worth listening for, not acting on yet: some said a chatbot felt easier to talk to than a person, and some said its answers were generic, wrong, or cut off when self-injury came up. Nothing here shows any tool helped or harmed, so ask your child what they use and what it does for them, and bring anything that worries you to a professional.

Who paid: The work was supported by a grant from the National Institute of Mental Health (K01MH131898); the article does not say whether the funder had any role, and it reports no equipment or software lent by others.

Do not take this as professional medical advice.

The longer versionThe passages copied from the paper, the pictures, and every source with how much of it we read · 1251 words · about 6 minRead it →Close

Teens Who Hurt Themselves Are Already Talking to ChatGPT. A New Study Asked Them What They Get—and What Backfires.

Twenty-one adolescents described using apps, online communities and AI chatbots to ride out self-injury urges. Nobody measured whether any of it worked.

How it could look · illustration generated by weeklyAI.watch, not a photograph

The article describes a study in which nonsuicidal self-injury—the term researchers use for hurting yourself without intending to die—is linked to a range of co-occurring mental health problems and to a higher risk of suicide¹. The article describes a study in which many adolescents who do it never receive formal treatment; stigma, cost and a shortage of providers all stand in the way, which is why researchers have been looking for alternatives that reach young people where they already are².

That "where they already are" is the subject of a new qualitative study published in Frontiers in Digital Health. Researchers at Northwestern University's Feinberg School of Medicine, the University of California Irvine and the nonprofit Mental Health America interviewed 21 adolescents aged 14 to 18, all of whom had self-injured on at least two days in the previous month³. The interviews ran from May 2024 to April 2025, were conducted remotely by a licensed clinical social worker, and were analyzed using thematic analysis—a method for finding patterns across open-ended conversations. The study was funded by a career grant from the National Institute of Mental Health.

The adolescents described self-injury as a coping strategy embedded in broader emotional and social stress—school pressure, family and cultural expectations, the pull of social media. They were aware of the harms, but many were ambivalent about stopping⁴. Stigma and fears about disclosure often kept them from seeking help, and that same reluctance pushed them toward digital tools for information and support⁵.

What they reported using is a familiar list: music streaming platforms, journaling and mood-tracking apps, online communities, and AI systems such as ChatGPT. They used them to regulate emotions, reflect on their experiences, seek advice, and discuss things they found hard to say out loud⁶. They valued the accessibility, privacy and immediacy. They also named the limits: inaccurate or generic AI responses, thin moderation in online communities, and the unpredictable emotional effects of music⁷.

The AI accounts were specific and, in the participants' telling, double-edged. One said talking to ChatGPT was "pretty helpful"⁸; another called it "my little online therapist" and said it felt more comfortable than confiding in someone they knew⁹. One explained, "It's almost easier to open up to, because you know it's not human"¹⁰; another said, "You can rant guilt free to a bot because you're not being annoying"¹¹. But safety filters frequently cut conversations short when self-injury came up—one participant said the system "doesn't want to continue that conversation, and sometimes I need to talk about it"¹²—and when the systems did respond, the advice often felt canned: "Confide in a friend or journal"¹³.

We could read only the summary of that study; the full text could not be fetched.

We could read only the summary; the full text could not be fetched.

The study's limits matter as much as its findings. Twenty-one adolescents is a small group, all recruited through a mental health nonprofit's website, which means they may have been more aware of their mental health and more receptive to digital tools than most teenagers³. The sample was mostly White and mostly female. Adolescents with severe mental health diagnoses or severe suicide risk were excluded, so the findings do not speak to higher-risk youth. And crucially, the interviews captured what adolescents said about their experiences—not measured outcomes. No app, platform or chatbot was tested for effectiveness or safety.

The study's authors also flag a timing problem: interviews ran over an 11-month period during which AI tools changed rapidly and public debate about AI and youth mental health shifted considerably. The specific chatbot behavior described may already differ from what a teenager encounters today. And the adolescents were often embarrassed to admit using AI at all—one said, "I talk to ChatGPT. I'll admit it, I talk to ChatGPT sometimes"⁸—which suggests the accounts may understate how much this is happening.

In this study, some adolescents described a chatbot that asked questions, remembered details and gave something back. They said it was easier to open up to, because it was not human. One participant said it was helpful when there was nobody else to talk to. In this study, the adolescents who used chatbots described not better advice, but a back-and-forth that felt mutual. In this study, adolescents described chatbots as something that asks and remembers, and said they kept using them even when the advice was generic. What you can do with this: ask what they get from talking to a chatbot that they don't get from a person, and listen for words like "it asks me questions" or "it remembers." That tells you whether the tool is filling a gap in human contact—which is the part you can actually act on.

In these interviews, adolescents who feared that a therapist or a parent would be told about their self-injury described keeping the most sensitive parts of their experience to apps and chatbots, and hiding or disguising those apps. In this study, participants described mandated-reporting rules and fears of parental notification as barriers to care, and asked for discreet app names and quick-exit screens²⁵. In this study, adolescents described talking more openly with an app than with the people around them about this subject, and taking steps to keep that use private. What you can do with this: before asking what your teenager tells an app, make clear in plain words who you would and would not have to tell, and when. A teenager who knows the rule is less likely to route everything around you.

The same apps that comfort can also trap. In this study, some adolescents described using social media to distract themselves from urges, then losing hours to scrolling and feeling worse about themselves. Some described online communities where self-injury became competitive—"If you see someone who does it more than you... you think, 'wow, I'm not doing it right'"²⁶—and one called these spaces "band-aids over a bullet wound"²⁷. In this study, adolescents described some apps as easy to put down and others as hard to put down. What you can do with this: ask which apps they can stop using when they want to and which ones use them. Then treat the second group as a design problem to solve together—timers, phone out of the room, a different playlist—rather than a character flaw to correct.

The study's own suggestions for future tools were concrete: one-tap access to coping strategies, mood and urge tracking with visual summaries, short-form journaling, curated music, and privacy features like biometric login, two-factor authentication, quick-exit screens and neutral cover interfaces²⁵. The authors also note that adolescents already have these tools in their daily routines, and that refining what exists may matter more than inventing something new.

What this study makes possible for you is not a verdict on any app. It is a set of questions you can ask tonight, and a way to listen to the answers. Some adolescents in this study said they talked to a chatbot about things they had not told other people. That is not a failure—it is information. The question worth carrying into the conversation is not "Is AI safe?" but "What does this tool give you that you are not getting from people—and what would it take for you to get more of that from us?"

Where each piece of context comes from, and how much of it we read

Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "Nonsuicidal self-injury (NSSI)—hereafter referred to as “self-injury”—is reported by approximately 17% of adolescents and is associated with a range of co-occurring mental health challenges and elevated risk of suicide ( 1 , 2 )."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "Yet, most adolescents that engage in self-injury do not receive formal treatment. Stigma surrounding mental health concerns, and self-injury in particular, as well as structural barriers such as treatment costs and limited access to providers, contribute to this gap and highlight a need for alternative forms of intervention."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "We conducted semi-structured interviews between May 2024 and April 2025 with 21 adolescents (14–18 years old) with lived experience of self-injury."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "Adolescents described self-injury as a coping strategy embedded within broader emotional and social stressors. While they were aware of the harms, they commonly expressed ambivalence about stopping the behavior."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "Stigma and concerns about disclosure often limited help-seeking and contributed to their use of digital technologies to access information and support."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "Participants reported using a range of tools, including music streaming platforms, journaling and mood-tracking apps, online communities, and AI systems such as ChatGPT to regulate emotions, reflect on their experiences, seek advice, and discuss sensitive topics."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "Although these technologies were valued for their accessibility, privacy, and immediacy, participants also highlighted limitations, including inaccurate or generic AI responses, limited moderation in online communities, and the complex emotional impacts of music."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "PID16 commented “I talk to ChatGPT. I'll admit it, I talk to ChatGPT sometimes. And honestly, that's pretty helpful for me”"
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "PID11 noted, “I message ChatGPT… it feels like it's my little online therapist, and that helps,” emphasizing that talking to an AI could feel more comfortable than confiding in someone they knew."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "PID16 explained, “It's almost easier to open up to, because you know it's not human.”"
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "PID12 shared this sentiment, “You can rant guilt free to a bot because you're not being annoying. But it's like, if it's another person, you can't talk to a person to this extent or expect them to be there to this extent.”"
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "PID11 described how safety restrictions often blocked important conversations they hoped to have: “I’ve used a few different AI therapy-specific models. But they’re like, a little bit more limited, or if I want to talk about maybe self-injury – it’ll just be like, you can't, we’re not talking about this – like, it doesn't want to continue that conversation, and sometimes I need to talk about it.”"
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "PID18 explained, “If I was having a particular struggle moment, a very intense moment, I wanted specific answers to my specific situation — [But] it was also like, ‘Confide in a friend or journal,’ and a lot of the other very general mental health reminders that I’ve been surrounded by in communities.”"
McBain RK, Cantor JH, Breslau J, Diliberti M, Zhang LA, Zhang F, et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015 — only the abstract - the full text could not be fetched — the passage: "19.2% of adolescents and young adults (population-weighted n = 8 207 180) in 2025 reported having used AI chatbots for mental health advice."
McBain RK, Cantor JH, Breslau J, Diliberti M, Zhang LA, Zhang F, et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015 — only the abstract - the full text could not be fetched — the passage: "Among those who sought advice from AI chatbots, 42.8% did so at least monthly, and 91.7% rated the advice as somewhat or very helpful."
McBain RK, Cantor JH, Breslau J, Diliberti M, Zhang LA, Zhang F, et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015 — only the abstract - the full text could not be fetched — the passage: "Most adolescents reported they had not disclosed AI chatbot use for mental health advice to anyone (63.3%)."
McBain RK, Cantor JH, Breslau J, Diliberti M, Zhang LA, Zhang F, et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015 — only the abstract - the full text could not be fetched — the passage: "respondents aged 18 to 21 years compared with respondents aged 12 to 14 years (aOR, 3.65; 95% CI, 1.98-6.74)"
McBain RK, Cantor JH, Breslau J, Diliberti M, Zhang LA, Zhang F, et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015 — only the abstract - the full text could not be fetched — the passage: "those who had spoken with a physician about their mental health in the prior 6 months compared with those who had not (aOR, 1.89; 95% CI, 1.18-3.03)"
McBain RK, Cantor JH, Breslau J, Diliberti M, Zhang LA, Zhang F, et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015 — only the abstract - the full text could not be fetched — the passage: "AI chatbots are already embedded in many youths' mental health information ecosystem, underscoring the need for parents and clinicians to proactively discuss chatbot use to promote safety, appro"
Maheux AJ, Boeldt D, Akre-Bhide S, Flannery J, Paige A, Villella G, et al. (2026). How US Youth Use AI Chatbots: Conversation Patterns From Naturalistic Keystroke Observations. Journal of Medical Internet Research. https://doi.org/10.2196/95819 — only the abstract - the full text could not be fetched — the passage: "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."
Maheux AJ, Boeldt D, Akre-Bhide S, Flannery J, Paige A, Villella G, et al. (2026). How US Youth Use AI Chatbots: Conversation Patterns From Naturalistic Keystroke Observations. Journal of Medical Internet Research. https://doi.org/10.2196/95819 — only the abstract - the full text could not be fetched — the passage: "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%)."
Maheux AJ, Boeldt D, Akre-Bhide S, Flannery J, Paige A, Villella G, et al. (2026). How US Youth Use AI Chatbots: Conversation Patterns From Naturalistic Keystroke Observations. Journal of Medical Internet Research. https://doi.org/10.2196/95819 — only the abstract - the full text could not be fetched — the passage: "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."
Maheux AJ, Boeldt D, Akre-Bhide S, Flannery J, Paige A, Villella G, et al. (2026). How US Youth Use AI Chatbots: Conversation Patterns From Naturalistic Keystroke Observations. Journal of Medical Internet Research. https://doi.org/10.2196/95819 — only the abstract - the full text could not be fetched — the passage: "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."
Maheux AJ, Boeldt D, Akre-Bhide S, Flannery J, Paige A, Villella G, et al. (2026). How US Youth Use AI Chatbots: Conversation Patterns From Naturalistic Keystroke Observations. Journal of Medical Internet Research. https://doi.org/10.2196/95819 — only the abstract - the full text could not be fetched — the passage: "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."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "Participants mentioned safeguards such as biometric login, two-factor authentication, quick-exit screens, or neutral “cover” interfaces may meaningfully increase uptake among adolescents who fear stigmatization."
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "PID16 said “A lot of issues like self-harm, especially online, can even be competitive in online spaces. If you see someone who does it more than you, or has more severe injuries than you, you think, ‘wow, I’m not doing it right,’ or ‘I can't even hurt myself good enough’”. She went on to say “It's just like a loop – you can make friends on there, but they’re also not focusing on recovery either. So, it's all of you gathered, just not recovering, and it becomes normal.”"
Ramirez, I., Barnas, O., Mohr, D. C. et al. (2026). Navigating self-injury in a digital world: adolescents’ perspectives on coping, help-seeking, and technology. Frontiers in Digital Health. https://doi.org/10.3389/fdgth.2026.1905943 - the article this story is about — the whole article — the passage: "PID21 reflected that visiting communities helped in the short term but had a negative effect in the long term: “Talking about it is keeping it from getting better, and encouraging other people to do it too. So I think things are just band-aids over a bullet wound.”"

Do not take this as professional medical advice.

The findings of other studies mentioned here are known to us through this document, which is the one we read; we did not open each of those studies.