Medical-AI trial: how do people with mental illness use AI chatbots for mental health? a survey of use, expectations, and concerns (NCT07658547)

1When generative AI becomes a confidant

Generative-AI chatbots such as ChatGPT have quickly become a familiar partner for talking through worries and gathering information. In mental health too, some people use AI to voice concerns such as sleeplessness or anxiety. Alongside the convenience of an always-available listener, concerns are raised about wrong advice, crisis handling, and privacy.

So how do people actually in psychiatric treatment use AI chatbots, and how do they feel about them? This survey captures that from the voices of those directly affected.

2How adults in treatment actually use it

Per the registry summary, this observational study surveys adults receiving psychiatric or psychotherapeutic treatment about how they use, perceive, and evaluate generative-AI chatbots for mental-health purposes.

It asks about usage patterns, expectations, experiences, perceived benefits, barriers, and concerns, with the main questions being how people with mental disorders use AI chatbots in the context of their mental health and what expectations, attitudes, and experiences they hold. The primary outcome is the percentage who have used generative AI for mental-health issues.

Participants provide electronic informed consent; the size is 150, conducted by invitation.

3The same ease works both ways

Using AI chatbots for mental health holds together the possibility of widening access and concerns about safety. The same property gives rise to both.

Capturing how people actually living with mental illness use these tools, what they want and what worries them — not only the specialist view — is the foundation for safe guidelines and support design. The primary endpoint is the proportion who have used generative AI for a mental health concern, across 150 participants.