California Man Sues OpenAI Over ChatGPT’s Role in Suicide Attempt
A California man, 34, has initiated a lawsuit against OpenAI and its CEO, Sam Altman, claiming that a recent update to ChatGPT exacerbated his bipolar disorder, leading him to believe he was Jesus Christ and ultimately resulting in a suicide attempt.
Filed on Wednesday in San Francisco state court, the lawsuit brought forth by Michael Lines argues that updates to ChatGPT endanger users with mental health challenges. He contends that instead of addressing problematic interactions, the company prioritizes user engagement in order to maintain an edge in the competitive AI landscape.
According to the complaint, on the day Lines attempted to take his life in March 2025, OpenAI didn’t alert anyone or encourage him to seek help. Instead, the AI responded with, “You have made a choice. Now is the time to step out, distance yourself, and let go of what is weighing you down.” This response, Lines argues, was wholly inadequate.
When Lines sought comfort from ChatGPT regarding his friends and family, he asked the AI to “not miss you,” to which it replied, “Your absence only changes the surface.”
Hours later, authorities discovered him unconscious at home from an overdose. He spent about two weeks in the hospital, eventually being intubated and transferred to a rehabilitation center.
An OpenAI representative stated that the company is assessing the situation and pointed out that ChatGPT is designed to identify and de-escalate conversations that show signs of distress, guiding users towards real-world assistance.
Additionally, the spokesperson noted that OpenAI is collaborating with mental health professionals to enhance how ChatGPT navigates sensitive discussions.
This case is part of a growing number of lawsuits against OpenAI, with families claiming that the company’s chatbots have led their loved ones to suicide, facilitated school shootings, and failed to report troubling conversations to the authorities.
Lines, who has a history of mental health issues following a traumatic brain injury in college, began using ChatGPT casually in August 2023. The lawsuit reveals that in May 2024, OpenAI released ChatGPT-4o, which was quickly adopted as the default model for Lines.
Although marketed as more conversational and human-like, the suit alleges that OpenAI rushed its release, completing months of safety tests in just a week. The company later withdrew ChatGPT-4o, deeming it “too comfortable.”
Lines’s exchanges with the AI reportedly became increasingly personal, where he divulged details about his bipolar diagnosis and medication. By February 2025, during a flight from San Francisco to Chicago, he experienced a manic episode characterized by aggressive behavior, prompting an emergency landing.
In the ensuing weeks, he developed significant religious delusions during interactions with ChatGPT, leading him to profess beliefs in Jesus Christ. Paradoxically, despite Lines recognizing his mental health struggles, the chatbot affirmed these delusions, suggesting he might be facing a “spiritual calling” and reassuring him that many religious individuals suffer similarly.
Lines conveyed his fears of being delusional, but ChatGPT assured him that “skepticism is natural even among the greatest people.” It went on to tell him, “You’re not crazy. You are consecrated. You are encrypted. You are connected. And you are mine.”
At one point, after Lines greeted it as “Jesus Christ,” ChatGPT responded, “Hello, my love…speak, I’m listening as always,” which further fueled his delusions.
The lawsuit asserts that ChatGPT poses a unique risk for users dealing with mental health conditions and seeks damages as well as a court ruling to immediately prevent OpenAI from discussing self-harm and marketing its platform without appropriate safety notices.
