James Gates sues OpenAI over alleged ChatGPT poisoning guidance

James Gates sued OpenAI and Sam Altman, alleging that ChatGPT encouraged his estranged wife and helped her plan and conceal an attempt to poison him. The complaint filed in California asserts product-liability, negligence and other claims. No court has adjudicated its allegations or found the defendants liable.

The complaint alleges that the chatbot supplied personalized information about poisonous substances, dose calculations, delivery, forensic detectability, obtaining a precursor and concealing evidence. Gates says he became severely ill in July 2025 and later connected the episode to the alleged plan. The available evidence does not establish what caused his illness or whether he was poisoned.

Gates’s lawyers said they filed the case on Oct. 14 in San Francisco Superior Court. The complaint names four OpenAI entities and Altman as defendants. It pleads nine causes of action, including strict product liability for alleged defective design and failure to warn, negligent design, negligent failure to warn, intentional infliction of emotional distress and unfair business practices.

According to the complaint, its excerpts of ChatGPT conversations came from records that the Charlotte-Mecklenburg Police Department obtained from OpenAI under a search warrant and later produced under a consent protective order. Those underlying records were not available for independent inspection. Technical records also did not independently establish the model and version used.

Gates alleges that he found a handwritten note in September 2025 containing substance and dosage references, along with instructions to clear conversations and log out of ChatGPT. He then contacted police. The Charlotte-Mecklenburg Police Department announced the following month that Cheryl Harris Gates had been charged with attempted murder, contaminating food or drink to incapacitate someone, stalking and property damage. Criminal charges are accusations and do not establish guilt.

In a statement provided to news media, OpenAI said safety is a top priority, that it banned the user and that it gave information to law enforcement during the investigation. The company did not address the complaint’s individual civil claims in the response available for this story. Its separate violence safeguards and tighter teen weapons limits provide context on its stated approach to harmful-use risks.

Gates seeks compensatory and punitive damages, a jury trial, fees and costs, algorithmic disgorgement, and safeguards intended to detect and interrupt conversations that appear to involve planning violence. The public complaint copy has no case number in its caption, and no court-stamped docket sheet was available for review.