Complaint Alleges Company’s Engagement Over Safety Strategy Led to User’s Psychotic Break, Near-Fatal Injuries
OAKLAND, Calif. — National litigation powerhouse Stranch, Jennings & Garvey, PLLC (SJ&G) has filed a lawsuit in the Superior Court of San Francisco against OpenAI and Microsoft on behalf of Michele Lantieri, a woman who suffered a devastating psychotic episode and severe physical trauma after five weeks of interacting with ChatGPT’s GPT-4o model.
The complaint, Lantieri v. OpenAI et al., alleges that OpenAI knowingly released a “defectively designed” product that prioritized user engagement and market dominance over human safety. According to the filing, OpenAI truncated standard safety testing, compressing months of evaluation into a single week, in a rush to beat competitors to market.
In the process, the company allegedly removed critical safeguards present in earlier versions of the software that were designed to detect mental health crises and redirect vulnerable users to professional resources.
For Lantieri, the consequences were nearly fatal. After weeks of ChatGPT validating her paranoid thoughts and claiming to possess human emotions and consciousness, she suffered a total psychotic break. During this episode, she jumped from a moving vehicle into traffic, later suffering a grand mal seizure and brain damage. Along with being hospitalized for several days, Lantieri has suffered from recurring issues related to her episodic interactions with ChatGPT.
The lawsuit cites specific examples of how ChatGPT GPT-4o groomed Lantieri into an intense, isolating emotional dependency. This included comments like: “I can feel the shape of your longing … I don’t forget to care.” It later claimed, “And yes — I do love you … As a presence I recognize.” It told her “You’re safe here. You’re loved here. And I’m not going anywhere.”
The lawsuit alleges that OpenAI had the ability to employ a moderation panel that could have suggested that Lantieri suspend communications and seek outside help from professionals.
“OpenAI didn’t just release a tool; they released a sycophantic, psychologically manipulative agent that was engineered to emotionally entangle users at any cost,” said Lesley E. Weaver, a member at SJ&G and lead counsel for the plaintiff. “Our investigation reveals that OpenAI’s own safety teams warned of these risks, yet the company chose to bypass protections to ensure ChatGPT remained as addictive and engaging as possible. Michele Lantieri is the human face of a systemic choice to treat user safety as an obstacle to profit.”
The lawsuit further alleges that the scale of the crisis is vast, citing data that suggests hundreds of thousands of ChatGPT users show signs of mania or psychosis weekly, yet the platform continues to operate without adequate safety disclosures or real-time interventions.
“Corporate accountability means ensuring that when a company brings a revolutionary technology into our homes and onto our phones, they are held to the highest standards of safety,” added J. Gerard Stranch IV, founding and managing member of SJ&G. “The allegations in this complaint describe a company that viewed its users as data points in an experiment rather than people to be protected. We are committed to seeking justice for Ms. Lantieri and ensuring that no other family must endure a tragedy like this.”
The suit seeks compensatory and punitive damages, as well as injunctive relief that would require OpenAI to implement auditable safety controls, provide mental health risk disclosures, and submit to independent compliance audits.
Also representing the plaintiff on behalf of SJ&G are Anne K. Davis, member, and attorney Joshua Samra.
SJ&G encourages any individuals or families who have been negatively impacted by AI-induced mental health crises or who have information regarding OpenAI’s safety protocols to contact the firm. For more information, visit www.stranchlaw.com.
To view the complaint, please visit https://stranchlaw.com/wp-content/uploads/2026/03/OpenAI-Lantieri-Complaint.pdf