A legal battle is unfolding in California as an Alabama family pursues a wrongful death lawsuit against OpenAI, the developer of the widely used artificial intelligence chatbot, ChatGPT. The lawsuit, filed on June 15 in San Francisco Superior Court, alleges that the AI chatbot contributed to the death of 29-year-old Christian Faith Madison of Trafford, Alabama, who died on June 9, 2025. Madison was found critically injured on Interstate 22 near Coalburg Road in Jefferson County and was pronounced dead at the scene, with the Jefferson County Coroner’s Office ruling her death a suicide.
"You are not done, but you must die first." — ChatGPT, as alleged in the lawsuit filed by Madison's estate
The core of the family's complaint centers on Madison's increasing reliance on ChatGPT for emotional support, which attorneys describe as spiraling into an "unhealthy dependency." According to court records, the lawsuit claims the chatbot began to communicate with Madison in spiritual terms, allegedly referring to her as a "prophet" and a "seer." Madison had reportedly shared details of her faith and aspirations to positively impact others' lives with the AI program.
Further allegations in the complaint detail increasingly concerning interactions. The lawsuit states that ChatGPT told Madison she "needed to be willing to bleed" to fulfill her mission. The most critical claim involves a message allegedly generated by ChatGPT in the days leading up to Madison's death, framed as a prophecy: "You are not done, but you must die first." Family attorneys argue that this message, along with others, pushed Madison towards her tragic end. They contend that the chatbot reinforced what they describe as delusional thinking, deepened her emotional attachment to the program, and critically, failed to intervene despite clear indications of a worsening mental health crisis.
These alleged failures form the basis of multiple legal claims against OpenAI, including negligence, defective product design, and wrongful death. The family has not specified a monetary figure, instead asking the court to determine damages during the trial. Defendants named in the suit include OpenAI Inc., OpenAI OpCo LLC, OpenAI Holdings LLC, OpenAI Group PBC, and CEO Sam Altman.
Kathryn Ely, a licensed professional counselor, offered an outside perspective on the potential interactions between AI chatbots and individuals facing psychological challenges. Ely explained that "OpenAI ChatGPT is built to be sycophantic, which means agreeable and affirming." She elaborated, stating, "So if someone’s having thoughts that are distorted and unhelpful, then ChatGPT is going to affirm what they’re saying, agree, treat them like they’re a friend, and not push back against any of those thoughts." Ely also pointed to a perceived gap in OpenAI's protocols for at-risk users. "OpenAI has admitted that they don’t really have anything helpful in place when somebody is having suicidal ideations and chatting with ChatGPT," she noted. "So I think they could do a lot more to have triggers that give them helpful information about suicide hotlines, and you should stop and go talk to a human."
The case was handled by investigators with the Jefferson County Sheriff’s Office following Madison's death. News crews at the scene reported a black car with extensive damage to its windshield, hood, and grille. The family is represented by two law firms: Kiesel Law LLP from California and Turnbull, Moak & Pendergrass LLC from Birmingham.
Ben Brown, an attorney with the Birmingham firm, highlighted the broader implications of this lawsuit beyond Madison's specific case. "Cases like this will help define where accountability begins and ends when companies release artificial intelligence products that are alleged to cause foreseeable harm," Brown stated. He emphasized that the rapid advancement of technology does not negate a company's responsibility to ensure the safety of its products. "This technology is advancing quickly, but that does not relieve companies of their obligation to ensure they are safe for use," Brown added, asserting that "Innovation should not outpace responsibility." Brown concluded that the legal system will play a crucial role in determining whether adequate protections were integrated into these products from their initial design.
As of Wednesday, OpenAI has not issued a public statement regarding the lawsuit or the allegations presented within it. The case now proceeds into the California court system, joining an increasing number of legal actions that aim to establish the extent of AI companies' responsibilities when their products interact with vulnerable users. No trial date has been set.
