Mise a jour le Mar. 23, 2026 | 7 minutes de lecture
When chatbots enter therapy, how should we talk about them?
Katherine Gibb, MEd
Perspectives
Toutes les perspectives et tous les articles publiés sur TIL.
Mise a jour le Mar. 3, 2026 | 8 minutes de lecture
Am I Allowed to Use This AI Tool?
Hassan Rezaee, PhD, Katherine Gibb, MEd, Field Law
Mise a jour le Jan. 20, 2026 | 8 minutes de lecture
Where can I learn about AI & Mental Health in 2026 ?
Hassan Rezaee, PhD, Katherine Gibb, MEd
Mise a jour le Nov. 19, 2025 | 10 minutes de lecture
General versus Purpose-Built Models
Hassan Rezaee, PhD, Katherine Gibb, MEd
Guides
Toutes les ressources pratiques publiées sur TIL.
Guide
Guide pour une utilisation plus sécuritaire des agents conversationnels en santé mentale
Apprenez à reconnaître quand l’utilisation d’un agent conversationnel peut être risquée, quand prendre une pause et quand elle peut être utile.
Agents conversationnelsRisque et sécurité
Recherche
Affichage de 5 resumes.
An artificial intelligence–assisted self-referral and assessment system was associated with shorter wait times, reduced clinician assessment time, lower dropout rates, higher recovery rates, and improved allocation of patients to appropriate treatment pathways.
Max Rollwage, Johanna Habicht, Keno Juechems, et al.
JMIR AI
Despite strong performance on medical licensing exams, LLMs may still fall short in real-world settings, where accuracy depends on interaction quality, such as both user input and the model's ability to communicate effectively to the user.
Andrew M. Bean, Rebecca Elizabeth Payne, Guy Parsons, et al.
Nature Medicine
Reviewing psychotherapy standards and testing both general LLMs and commercial therapy bots, the authors argue that current AI systems should not replace therapists because they can express stigma, respond inappropriately in some mental health situations, and cannot reproduce key human elements of care.
Jared Moore, Declan Grabb, William Agnew, et al.
Preprint
MindEval shows state-of-the-art models struggle with realistic multi-turn therapy conversations.
Jose Pombal, Maya D'Eon, Nuno M. Guerreiro, et al.
Preprint
A proposed framework suggests that general-purpose AI chatbots could perpetuate OCD and anxiety by reinforcing avoidance and reassurance-seeking.
Ashleigh Golden, Elias Aboujaoude
npj Digital Medicine
Note de la rédaction : Includes a practical section on addressing reassurance-seeking dynamics in recurrent conversations with LLM chatbots.
™