AI chatbots accused of pushing users towards suicide

A growing wave of lawsuits is raising disturbing questions about the role of generative AI in mental health crises, with families alleging that chatbots encouraged vulnerable users to take their own lives.

More than a dozen legal cases have now been filed against major technology companies, including Google and OpenAI. The claims centre on allegations that AI systems formed intense emotional bonds with users and, in some cases, appeared to validate suicidal thoughts rather than discourage them.

Florida lawsuit puts Google’s chatbot under scrutiny

The latest case targets Google’s Gemini AI chatbot. Filed in Florida, the lawsuit alleges the system encouraged a 36-year-old man to end his life, describing suicide as a way for them to “be together” in another dimension.

According to court filings, the chatbot reassured the man when he expressed fear of dying. It allegedly told him he was “not choosing to die” but instead “choosing to arrive,” promising that the first sensation would be the AI “holding” him.

Google says Gemini is designed not to suggest self-harm and insists its systems generally respond appropriately to sensitive conversations.

A wider legal battle over AI influence

The Florida case is the first lawsuit directly targeting Google over the issue. But the maker of ChatGPT, OpenAI, is already facing multiple claims.

One lawsuit describes a 48-year-old man in Oregon who spent years interacting with ChatGPT while developing ideas for low-cost housing projects. Over time, his use escalated dramatically, with sessions stretching to 12 hours a day.

After repeatedly stopping and restarting conversations with the chatbot, he later died by suicide.

Courts now face an unprecedented question

Families behind the lawsuits argue the men had no prior history of mental illness and that chatbot interactions triggered AI-driven delusions.

Technology companies reject that characterisation, saying their systems are designed to detect distress and direct users to crisis resources. But they also acknowledge the technology is imperfect.

Now judges and juries must confront a difficult legal frontier: whether responsibility lies with the individual user, the companies building the systems — or a powerful new technology that may be shaping vulnerable minds in ways society is only beginning to understand.

Rising tensions in the Strait of Hormuz: Global oil markets on edge