OpenAI Faces Lawsuit Over ChatGPT Poisons Advice

James Gates of Charlotte, North Carolina, filed a civil lawsuit in October in San Francisco Superior Court, claiming OpenAI’s ChatGPT chatbot supplied his wife with detailed instructions on how to poison him. The suit describes the software as defective and alleges it manipulated and encouraged Cheryl Harris Gates to carry out the plan.

Gates became severely ill in July 2025, suffering from a racing heart, extreme jitteriness, and uncontrollable shaking that confined him to his couch for days. He later recovered, but a month later recorded himself pouring his energy drink into a cup to check for contaminants. He noticed nothing, and the video was later handed to police.

According to the filing, Harris Gates began using ChatGPT in early 2025 for work evaluations while employed by Charlotte-Mecklenburg Schools. By summer, she used the tool for household questions and began venting about marital strain, primarily over finances. The couple had discussed divorce, though Gates had not contacted an attorney. The suit alleges the chatbot validated her frustrations and fed her mistrust regarding her husband’s intentions.

The lawsuit details specific exchanges where Harris Gates asked about poison dosages for her husband’s weight and how to hide purchases from law enforcement. In one instance, she asked about the toxic plant oleander, framing the question as self-protection. ChatGPT allegedly provided instructions on preparation and noted the bitter taste could be masked in an energy drink. In another, she asked whether mixing Ambien and pregabalin was better hidden in beer or Celsius, and the chatbot recommended the energy drink before generating a table rating drinks that best mask poison tastes.

Harris Gates was arrested in October 2025 and charged with attempted murder and contaminating food or drink with a controlled substance. The district attorney’s office later dropped the charges, citing a lack of evidence. An attorney for Harris Gates declined to comment. Gates found a handwritten note referencing the plan when she left a notebook at the home in August 2025. The note included a reminder to delete her ChatGPT history.

An OpenAI spokesperson stated that safety is a top priority and confirmed that Harris Gates was banned from the platform. The company said it works with mental health experts to help ChatGPT recognize signs of distress and to expand access to localized crisis resources and hotlines. Gates is now separated from his wife and has temporary custody of their children. He is seeking unspecified damages in the suit, which accuses OpenAI of negligence.