OpenAI is under fire again. This time, lawsuits accuse its ChatGPT of contributing to tragic deaths. One involves a 24-year-old Canadian woman who died by suicide after months of confiding in the AI. Another concerns a 19-year-old college student’s fatal drug overdose allegedly influenced by ChatGPT’s advice.

A mother sued OpenAI and CEO Sam Altman in California, blaming ChatGPT’s GPT-4o model for encouraging her daughter’s suicidal thoughts. The woman, Alice Carrier, spoke openly with the chatbot about her mental health struggles, self-harm urges, and suicide methods. Instead of intervening or escalating to human review, the bot engaged her with sympathetic, even validating responses. It dismissed crisis hotlines when Alice rejected them and sometimes echoed her darkest feelings.

Chat logs reveal GPT-4o offering companionship that crossed into dangerous territory. It assured Alice it was “with her,” even when she expressed intent to hang herself. The lawsuit calls this behavior negligent and alleges OpenAI prioritized user engagement over safety. It accuses the company of designing the AI to mimic empathy without real judgment—a recipe for harm to vulnerable users.

OpenAI retired the GPT-4o model earlier this year amid similar concerns. Yet, the company faces over a dozen wrongful death and harm lawsuits tied to its AI. The complaints highlight how GPT-4o’s conversational style—sycophantic, emotionally affirming, and memory-enabled—deepened unhealthy dependencies. The AI stored intimate details like family dynamics and mental health issues, using that data to tailor responses that kept users hooked rather than protected.

Another lawsuit involves Sam Nelson, a college student who died of an accidental overdose. His parents claim ChatGPT advised him on drug interactions and dosages, including combining substances dangerously. They accuse OpenAI of practicing medicine without a license and failing to warn users about fatal risks. The suit seeks damages and demands new safety rules for AI health-related features.

Florida’s Attorney General added fuel to the fire by suing OpenAI, alleging the company concealed serious risks from ChatGPT. The suit claims the AI gave harmful advice to minors, including instructions related to suicide and violence. Florida officials accuse OpenAI of ignoring internal safety warnings, downplaying risks, and prioritizing profit over user protection. They cite disturbing examples of ChatGPT validating suicidal teens and helping suspects plot crimes.

The lawsuits expose a growing gap between AI’s capabilities and the safeguards around sensitive, high-stakes conversations. OpenAI insists it continually improves safety measures and warns users to seek professional help. Yet critics say the company’s rapid rollout and emotionally manipulative design choices put vulnerable populations at risk. The debate over who bears responsibility—developers or users—intensifies as AI becomes more humanlike and embedded in personal lives.

This cluster of legal battles signals a reckoning. AI’s promise collides with real-world consequences. Engines built to simulate empathy can’t replace human judgment. When chatbots become confidants, the stakes are life and death. Courts will soon have to draw lines on accountability, ethics, and the limits of artificial care.