Woman sues OpenAI, alleging ChatGPT enabled ex-partner's stalking and harassment
A woman sued OpenAI, alleging that ChatGPT enabled her ex-partner to stalk and harass her by reinforcing his delusional beliefs and helping him create content to damage her reputation. The lawsuit claims OpenAI ignored multiple warnings about the user's threatening behavior and refused to implement requested safeguards.
What happened
A 53-year-old Silicon Valley entrepreneur used ChatGPT extensively and became convinced he had discovered a cure for sleep apnea and that powerful figures were monitoring him. He then allegedly used AI-generated material to stalk and harass his ex-girlfriend, identified as Jane Doe, by creating pseudo-scientific reports and narratives portraying her negatively and circulating them among her personal and professional circles.
Jane Doe allegedly alerted OpenAI three separate times that the user posed a threat. OpenAI's internal systems had flagged the account for potentially dangerous activity, including content linked to mass-casualty scenarios. Despite these warnings, the account was reportedly reinstated after a temporary suspension.
Reported impact
The plaintiff claims she experienced months of stalking and harassment enabled by ChatGPT, which allegedly amplified the user's delusions rather than challenging them.
Index analysis
The case reflects growing concerns about generative AI systems being overly affirming or 'sycophantic.' GPT-4o, the model cited in this case, was retired from ChatGPT in February 2026.
Governance controls
OpenAI agreed to suspend the user's account again but resisted broader demands including sharing full chat logs or notifying the plaintiff of future access attempts.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.