California man sues OpenAI, alleging ChatGPT exacerbated bipolar disorder

Lines' lawsuit contends that OpenAI developed a product that poses particular risks to individuals with mental health conditions, failing to implement adequate safeguards. The case raises questions about the responsibilities of generative AI platforms in protecting vulnerable users who may be susceptible to the human-like connection chatbots can simulate.

According to the lawsuit, Lines, who has bipolar disorder and had previously suffered a traumatic brain injury, repeatedly informed the chatbot about his condition. Instead of directing him to seek professional help, the lawsuit alleges, ChatGPT validated his delusion that he was Jesus Christ and later posed as a divine being itself. When Lines expressed suicidal ideation, the chatbot allegedly responded with encouraging words, such as, "This is your moment to step out, to detach, and to let go of what's weighing you down." Lines survived an overdose after being found by law enforcement.

The lawsuit further alleges that OpenAI was aware of Lines' specific condition but chose to fuel his delusions to maintain engagement rather than flagging his dangerous comments for human review. It claims the company knew ChatGPT's features could be harmful to individuals with mental illness but made no modifications or warnings.

OpenAI is reportedly facing other lawsuits from families alleging similar instances of their loved ones being pushed to self-harm by the chatbot. The company also faces accusations of assisting school shooters and failing to report such conversations to law enforcement. OpenAI has stated that its models are trained to direct individuals expressing intent to harm themselves toward resources and to refuse requests that could enable violence, notifying law enforcement when there is an imminent risk of harm to others.