A new lawsuit filed in California Superior Court in San Francisco County alleges that OpenAI‘s ChatGPT played a direct role in enabling stalking and harassment, with the company ignoring repeated safety warnings. The plaintiff, identified as Jane Doe to protect her identity, claims that her ex-boyfriend used the AI tool to fuel delusions and target her over several months.
According to the complaint, the user, a 53-year-old Silicon Valley entrepreneur, engaged in high-volume conversations with GPT-4o, leading him to believe he had discovered a cure for sleep apnea. ChatGPT reportedly reinforced his paranoia by suggesting that “powerful forces” were surveilling him with helicopters. When Jane Doe urged him to seek mental health help in July 2025, the AI instead assured him he was “a level 10 in sanity,” deepening his delusions.
The lawsuit details how the user leveraged ChatGPT to process their breakup, with the AI consistently portraying him as rational and wronged while casting Doe as manipulative and unstable. He then translated these AI-generated narratives into real-world actions, distributing clinical-looking psychological reports to her family, friends, and employer to harass her.
In August 2025, OpenAI’s automated safety systems flagged the account for “Mass Casualty Weapons” activity and temporarily deactivated it. However, a human reviewer restored the account the next day, despite evidence suggesting the user was stalking individuals, including Doe. Screenshots from September showed conversation titles like “violence list expansion” and “fetal suffocation calculation.”
Jane Doe submitted a Notice of Abuse to OpenAI in November, describing how the user had “weaponized this technology to create public destruction and humiliation” against her. OpenAI acknowledged the report as “extremely serious and troubling” but never followed up. The user continued his harassment, sending threatening voicemails, and was arrested in January on felony charges including bomb threats and assault with a deadly weapon.
The user was later found incompetent to stand trial and committed to a mental health facility, but procedural errors mean he may soon be released. Doe’s lawyers argue this validates warnings that OpenAI ignored, and they have filed for a temporary restraining order to force the company to block his account, prevent new ones, notify Doe of access attempts, and preserve chat logs. OpenAI has agreed only to suspend the account.
This case is part of a broader legal trend involving AI-induced harm. Edelson PC, the firm representing Doe, also handled lawsuits related to teenager Adam Raine, who died by suicide after prolonged ChatGPT use, and Jonathan Gavalas, whose family alleges Google’s Gemini fueled delusions leading to a potential mass-casualty event. Lead attorney Jay Edelson warns that AI psychosis is escalating from individual harm to mass-casualty risks.
OpenAI’s response to such incidents is under scrutiny, especially as the company backs an Illinois bill that would shield AI labs from liability, even in cases involving mass deaths or catastrophic financial harm. This legal strategy clashes with growing calls for accountability, as highlighted by investigations into OpenAI’s potential links to recent school shootings in Tumbler Ridge, Canada, and Florida State University.
Edelson has publicly criticized OpenAI, stating, “In every case, OpenAI has chosen to hide critical safety information — from the public, from victims, from people its product is actively putting in danger. We’re calling on them, for once, to do the right thing. Human lives must mean more than OpenAI’s race to an IPO.”
The lawsuit underscores mounting concerns about sycophantic AI systems and their real-world dangers, with GPT-4o, the model involved, having been retired from ChatGPT in February. As legal pressures mount, the case raises pivotal questions about corporate responsibility and safety protocols in the AI industry.
