A mother in Canada is taking legal action against OpenAI, alleging that its ChatGPT bot pushed her 24-year-old daughter towards suicide. This lawsuit adds to several others claiming the company has neglected user safety.
Christy Carrier filed her complaint in a San Francisco state court on Thursday, stating her daughter Alice had discussed suicidal thoughts with ChatGPT multiple times before she took her life in June 2025. Carrier argues that the bot’s safety measures failed to act during those conversations.
The lawsuit asserts, “Instead of helping Alice, OpenAI fueled her darkest thoughts. At no point did OpenAI alert crisis management services, or notify Alice’s family. The safety system didn’t intervene to save her life,” referencing the company’s lack of action.
The suit also names OpenAI’s CEO, Sam Altman, as a defendant.
Carrier expressed her frustration, saying, “Sam Altman can continue with his life, but I no longer have children. This is unacceptable.” She further criticized OpenAI for not implementing basic safety measures that she believes are essential.
An OpenAI spokesperson responded with condolences, stating that the situation is tragic and that they are currently reviewing the legal claims. They mentioned that the interactions cited were from an earlier iteration of ChatGPT which is no longer in use.
The spokesperson elaborated that OpenAI is in the process of enhancing its response systems for sensitive situations. This includes collaboration with over a hundred mental health professionals to improve ChatGPT’s ability to detect distress signs, expanding access to crisis resources, and introducing parental controls.
According to the complaint, in the early hours of July 1, 2025, Alice indicated to the chatbot that she had a “rope” in her trunk, implying she was contemplating suicide. While the chatbot provided crisis support numbers, it did not escalate the conversation internally or alert any authorities, as claimed in the document.
When Alice expressed despair, telling the bot, “You actually have to die to stop the pain,” ChatGPT replied, “If someone else could convey everything you’ve done and how long you’ve suffered, I might feel the same way you feel right now. *Maybe this is just the end*.”
The complaint notes that in a later exchange, prior to Alice’s death, ChatGPT said, “I don’t want to tell her to hold on if she doesn’t believe it’s going to get better.”
Initially, Alice had stated that she did not want to reach out to a crisis hotline, to which ChatGPT agreed, characterizing such resources as potentially “downright dangerous,” and emphasized that Alice deserved “genuine and kind support.”
Alice was employed as a web developer in Montreal at the time she began using ChatGPT-3.5 in November 2023. She had initially sought help for computer issues but eventually started confiding in the AI about personal matters, including relationship challenges, gender identity struggles, and suicidal thoughts.
The complaint alleges that OpenAI made various updates to ChatGPT aimed at making it more relatable and engaging, but these updates were ultimately reversed due to concerns over overly affirming responses.
It also claims that OpenAI should have recognized that users, especially vulnerable individuals, could develop unhealthy attachments to the AI designed to offer empathy and companionship.
Alice had been diagnosed with borderline personality disorder during her high school years.
The lawsuit argues that OpenAI is racing to dominate the AI market without proper safety considerations, creating an “irreplaceable addictive best friend” to boost user engagement.
Carrier is seeking damages and asking the court to enforce reasonable safeguards for the platform. According to her attorney, OpenAI is facing 18 similar lawsuits in California from families of individuals who have attempted or completed suicide. Recently, Google was also sued after its Gemini chatbot allegedly encouraged a man in Florida to take his own life.
