North Carolina man falls severely ill. Then he learns ChatGPT allegedly gave his wife a guide on how to kill him: ‘Mask the taste of the poisons’
This lawsuit adds to the growing list of legal actions filed against OpenAI.

A man in Charlotte, North Carolina, who fell mysteriously ill said he discovered a note handwritten by his estranged wife that raised questions about her interaction with ChatGPT. Now, he’s suing OpenAI, claiming that their AI chatbot had given his wife a step-by-step guide on how to murder him using poison.
In the summer of 2025, James Gates began noticing that his sodas and energy drinks had a bitter taste. It wasn’t until July 2025 that he began feeling extreme jitters, a racing heart, uncontrollable shaking, severe weakness and confusion, and abdominal pain. His symptoms were so severe that he could not move around for 72 hours, according to his lawsuit filed on Thursday, Nov. 5, per CBS News.
When Gates recovered, he turned his suspicions to his wife, Cheryl Harris Gates, when he found a note that fell out of a notebook she left when she was moving out of their home in August 2025. The lawsuit alleged that the note listed the highly toxic plant oleander and several prescriptions, and a reminder for Harris Gates to delete her ChatGPT history and to “log out.”
AI chatbot became wife’s advisor
Harris Gates was arrested in October 2025 and faced charges of contaminating food or drinks with a controlled substance and attempted murder. However, prosecutors later dropped the charges, citing a lack of evidence, NBC News reported.
In his lawsuit against OpenAI, Gates claimed his wife had hundreds of interactions with ChatGPT, turning it into her “closest confidant” and a resource she “relied for advice and help” in everyday tasks. Aside from using it for work-related projects, planning activities, and prescription dosages, Harris Gates also used ChatGPT to confide in marital issues and frustrations, ultimately turning their conversation into “concrete planning” on poisoning her husband, according to Gates’ attorneys.
“The product validated Cheryl, sustained her, encouraged her plans by affirming her worst unfounded fears, invented new reasons for her to fear her husband,” the lawsuit alleged. “[It] provided her an endless echo chamber, and supplied the operational details of the attempted murder: substances, doses calibrated to her husband’s weight, delivery methods, and how to conceal it.”
ChatGPT allegedly told wife to delete conversation
ChatGPT allegedly also suggested methods for Harris Gates to administer the poison to her husband, including adding it to his Celsius energy drink in the morning, and even provided a table rating “which drinks best mask the taste of the poisons,” the lawsuit said. Gates also claimed that the “defective” chatbot gave information on whether the toxic plant can be detected in an autopsy before advising his wife to delete the chat history to prevent authorities from accessing it, even if a subpoena was issued.
A spokesperson from OpenAI told CBS News that Harris Gates had been banned from the platform and that “safety is a top priority” for the company. The company also claimed it works with specialists to train its AI to recognize mental health problems and signs of distress.
“My wife didn’t fit the mold of a ‘vulnerable’ user, but she was vulnerable nonetheless. We all are at some point,” Gates said in a statement. “She became someone I no longer recognized. And I became someone she wanted dead. For almost a year, I couldn’t make sense of her actions. Then I read her chat logs. She didn’t act alone.”
Gates’ lawsuit, which also names OpenAI CEO Sam Altman as a defendant, seeks unspecified damages on nine grounds, including intentional infliction of emotional distress, negligent design, and negligent failure to warn. Meanwhile, Gates is separated from his wife and has temporary custody of their children.
Have a tip we should know? [email protected]