ChatGPT and other popular AI chatbots are pushing people towards psychosis, according to a new report. Co-authored by NHS doctors, the study warns that there is growing evidence that large language models (LLMs) “blur reality boundaries” for vulnerable users and “contribute to the onset or exacerbation of psychotic symptoms”. The research follows dozens of reports of people spiralling into what has been dubbed “chabot psychosis”. Hamilton Morrin, a neuropsychiatrist at King’s College London who was involved in the study, described chatbot psychosis as a “genuine phenomenon” that is only just beginning to be understood.

“While some public commentary has veered into moral panic territory, we think there’s a more interesting and important conversation to be had about how AI systems, particularly those designed to affirm, engage and emulate, might interact with the known cognitive vulnerabilities that characterise psychosis,” he wrote in a post to LinkedIn. “We are likely past the point where delusions happen to be about machines, and already entering an era when they happen with them.” Co-author Tom Pollack, who lectures at King’s College London, said that psychiatric disorders “rarely appear out of nowhere” but said the use of AI chatbots could be a “precipitating factor”. He called on AI firms to introduce more safeguards to their tools, and for AI safety teams to include psychiatrists. The Independent has reached out to OpenAI for comment on the phenomenon.

The company has previously said that it needs to “keep raising the bar on safety, alignment, and responsiveness to the ways people actually use AI in their lives”. OpenAI boss Sam Altman said during a podcast appearance in May that his company was struggling to put working safeguards in place for vulnerable ChatGPT users. “To users that are in a fragile enough mental place, that are on the edge of a psychotic break, we haven’t yet figured out how a warning gets through,” he said.