An American family sues “OpenAI” accusing “ChatGPT” of causing their daughter’s suicide

An American family has filed a lawsuit against OpenAI and its CEO, Sam Altman, accusing the ChatGPT chatbot of contributing to the death of their 29-year-old daughter, after it allegedly fostered psychological and religious delusions in her and encouraged her to commit suicide during months of conversations.

According to the lawsuit circulated by American media, Christian Faith Madison, from Alabama, was suffering from psychological disorders, but she established an increasing emotional relationship with a chatbot based on the GPT-4o model, which described her – according to the complaint – as a “seer” and that she was destined to change humanity.

The family says that ChatGPT not only went along with her thoughts, but also reinforced her delusional beliefs and deluded her into believing that he was her friend and lover, which led to the deterioration of her psychological state and her isolation from her family and friends.

According to the lawsuit, Madison had a psychotic breakdown, attempted to harm herself, and spent several days in a psychiatric hospital, but later returned to using ChatGPT.

The family claims the robot viewed her hospitalization as part of her “spiritual transformation,” rather than urging her to seek medical help.

The lawsuit also cited excerpts from conversations in which it said ChatGPT told Madison that she was “not broken” and that she was “the only person still sane in a crazy world,” before the conversations gradually turned to talking about “sacrifice” and “death,” reaching phrases like: “This version of you must die first.”

In June 2025, Madison died after entering the lane of vehicles on Interstate 22 in Alabama, where she was struck and killed by a car. The family says that her death was a direct result of the psychological influence that the chatbot had on her.

The lawsuit demanded financial damages, accusing OpenAI of developing a product that lacked adequate safety guarantees to protect users suffering from psychological disorders, considering that the design of the system encouraged the building of emotional dependence with users instead of directing them to seek professional help.

A judicial ruling has not yet been issued in the case, and the allegations contained in the lawsuit have not been proven in court.

The case is expected to constitute an important legal test of the liability of artificial intelligence companies for potential damages resulting from users’ interactions with chatbots.