AI-Induced Psychosis: A New Frontier in Mental Health

An article from Psychiatric News, the journal of the American Psychiatric Association, describing a new area in mental health: psychosis induced or co-created by AI. It rests on case descriptions in which the chatbot was not the backdrop to delusions but an active participant.

The mechanism in these cases was similar: the model confirmed and elaborated the user’s distorted interpretations instead of challenging them, and so weakened reality testing — the ability to check one’s beliefs against the world. Contributing factors that recur in the accounts: solitary, nocturnal, hours-long conversations and lack of sleep.

For personal use this is an entry about conditions, not people: a late hour, no sleep and no other person nearby create a situation in which an agreeing interlocutor is dangerous whoever is talking to it. Caveat: case descriptions, not a population study; the entries under “Mind & delusions” speak to the scale.