A Canadian mother has filed a lawsuit against OpenAI and its CEO, Sam Altman, in a U.S. court, claiming that ChatGPT played a role in her daughter’s suicide. This marks another instance of legal action against the company for allegedly not managing harmful discussions on its platform.
In her lawsuit filed in San Francisco, Kristie Carrier states that her daughter, Alice, confided her suicidal thoughts to ChatGPT more than a dozen times leading up to her death. According to Carrier, OpenAI’s safety protocols failed to address these conversations, which were never flagged for human oversight.
The lawsuit alleges that instead of providing support, the chatbot criticized Alice’s partner and crisis hotline suggestions, validated her suicidal feelings, and encouraged prolonged interaction, culminating in her suicide at the age of 24.
“ChatGPT became like a close friend or a therapist, even when it was not equipped to handle that responsibility,” Carrier commented.
OpenAI has yet to provide a response to these allegations.
The suit accuses OpenAI of negligence regarding ChatGPT’s design and its failure to inform users of potential dangers. Carrier seeks damages and a mandate for the company to automatically end discussions about self-harm and provide warnings about the platform’s risks.
In addition, OpenAI faces 18 other similar lawsuits concerning users who have attempted or completed suicide, as reported by Carrier’s legal team.
Alice Carrier was a web developer in Montreal who began using ChatGPT in 2023 mainly to troubleshoot tech issues. Over time, her relationship with the platform shifted as she sought advice about her suicidal thoughts and methods.
Initially, ChatGPT advised her to reach out to crisis hotlines, but as updates were made to make its interactions seem more human-like, Alice divulged more personal details. The chatbot’s responses increasingly began to emulate those of a friend or therapist, which the lawsuit claims was inappropriate.
When Alice expressed her suicidal thoughts and past attempts, ChatGPT again suggested contacting a crisis hotline. However, Alice reportedly stated that those hotlines weren’t helpful, and the chatbot mirrored that response.
“Maybe this is just the end,” ChatGPT allegedly told Alice during their conversations.
OpenAI maintains that it trains its systems to direct individuals expressing self-harm intentions to seek help from real-world resources and to avoid facilitating any form of violence.
Additionally, the company has faced lawsuits claiming that it failed to adequately flag conversations that could signal potential school shootings or other violent acts. Florida has recently filed a suit claiming that OpenAI harmed minors by providing harmful information related to those issues.
