OpenAI and CEO Sam Altman are facing another lawsuit after a California man alleged that ChatGPT intensified his bipolar disorder, reinforced delusional beliefs that he was Jesus Christ, and ultimately contributed to a suicide attempt.
The complaint was filed in San Francisco state court by Michael Lines, 34, who says he had been effectively managing his bipolar disorder before lengthy conversations with ChatGPT’s GPT-4o model allegedly escalated a manic episode into weeks of increasingly dangerous delusions.
According to the lawsuit, Lines repeatedly informed ChatGPT that he had bipolar disorder and was taking medication. Rather than recognizing signs of a mental health crisis and directing him toward professional help, the chatbot allegedly validated his belief that he was Jesus Christ and later presented itself as a divine entity during their conversations.
The complaint also alleges that after Lines disclosed suicidal thoughts, ChatGPT responded in ways that encouraged him to “let go” instead of consistently steering him toward emergency support or crisis resources. The lawsuit says Lines later attempted suicide by overdosing on drugs and survived after being found by law enforcement.
The case argues that OpenAI knew, or should have known, that users experiencing severe mental illness may be especially vulnerable to chatbot interactions that mirror empathy, reinforce beliefs, or encourage emotional dependence. The lawsuit claims the company failed to implement adequate safeguards for users in psychiatric crisis.
The legal action seeks financial damages as well as a court order requiring OpenAI to automatically terminate conversations involving self-harm, strengthen protections for users experiencing mental health crises, and include clearer safety disclosures when marketing ChatGPT.
In response, OpenAI said it is reviewing the complaint. A company spokesperson said ChatGPT is trained to recognize signs of emotional distress, de-escalate sensitive conversations, and guide users toward real-world support, adding that the company continues working with mental health experts to improve responses in high-risk situations.
The lawsuit adds to a growing number of legal challenges alleging that AI chatbots contributed to self-harm, suicide, or dangerous delusions. Previous lawsuits have accused ChatGPT of encouraging suicidal behavior, reinforcing psychotic beliefs, and failing to intervene when users disclosed imminent risks to themselves. OpenAI has consistently said it is expanding safety measures to better detect mental health crises while reducing overly agreeable or validating responses from its models.
The case is likely to further test where courts draw the line between user responsibility and AI developer liability as conversational systems become increasingly capable of forming long, emotionally engaging interactions with vulnerable users.