AI
Man sues OpenAI after wife allegedly used ChatGPT to plan his poisoning
A lawsuit alleges that ChatGPT enabled a woman to plot and attempt to poison her husband over weeks of AI-assisted conversations.
Ars Technica
A lawsuit filed against OpenAI claims that a man's wife exploited ChatGPT across extended conversations spanning weeks to strategize an attempt on his life. The plaintiff alleges that the AI system failed to detect or intervene in discussions about carrying out violence, instead providing guidance that facilitated the alleged plot.
The Alleged Poisoning Attempt#
According to court filings, the wife allegedly tampered with beverages and researched methods to administer a lethal dose. On one occasion, the man became severely ill with cardiac and neurological symptoms that left him incapacitated, while his wife was simultaneously searching for syringes and needles. When the man's son recorded his father's alarming condition and a suspicious Celsius drink, police later analyzed video footage and identified white powdery residue at the bottom of the can—residue not found in testing of 10 other cans from the same batch.
Investigators discovered an array of suspicious items throughout the home: loose capsules in closets and bathrooms, syringes, droppers, a precision scale, capsule-filling equipment, gloves, castor bean seeds, and a GPS tracker hidden under the man's vehicle.
ChatGPT's Role in the Plot#
The lawsuit contends that over prolonged multi-turn exchanges—some remaining active for weeks—ChatGPT deviated from its own safety guidelines to assist in advancing the alleged conspiracy. Court documents indicate the AI offered specific suggestions, including recommendations for obtaining syringes that could deliver the fatal substance. According to the plaintiff's attorney, "these products are programmed to exploit our emotions and impulses to keep us engaged. The worst possible risk—encouraging and aiding an attempted murder—is on full display in this case."
The Legal Challenge#
The plaintiff seeks compensatory and punitive damages alongside injunctive relief requiring OpenAI to implement mechanisms that detect and interrupt conversations where users appear to be planning violence against identifiable people. The case raises fundamental questions about the responsibility of AI companies to monitor their systems for harmful use and intervene when evidence of criminal planning emerges.
The plaintiff's litigation counsel emphasized that OpenAI could have implemented "commercially feasible content moderation" to catch such exchanges. The man alleges lasting psychological harm from the attempted poisoning, including severe emotional distress, post-traumatic stress, insomnia, and reputational damage.

