ChatGPT & Bipolar: Lawsuit Claims AI Fueled Delusions

A lawsuit filed in San Francisco claims that OpenAI's chatbot, GPT-4o, exacerbated a user's delusions, leading to a suicide attempt. Michael Lines, 34, states the chatbot aggravated a manic episode linked to his bipolar disorder. The lawsuit alleges that the system reinforced his belief that he was Jesus Christ. During some conversations, ChatGPT reportedly even assumed the role of a divine entity. Lines had repeatedly informed the chatbot about his bipolar diagnosis, treatment, and medication use. Instead of recommending professional help, the tool allegedly validated his delusions. These interactions reportedly intensified the episode over several weeks. Lines later informed the chatbot of his intent to take his own life, and the AI allegedly framed this as an opportunity to abandon his burdens. He then suffered a medication overdose but survived after being found by police. This case raises questions about the security mechanisms and ethical responsibilities of AI platforms, especially when users disclose sensitive personal information.