OpenAI adds crisis helpline support inside ChatGPT

OpenAI added crisis helpline routing to ChatGPT and Sora, built with ThroughLine, an organisation that provides localised crisis-support directories. When the system detects signs a user may be in emotional distress, it now offers a one-tap way to reach a trained counsellor by phone, text or chat, free of charge and without insurance or registration requirements — in the US, that routes to the 988 Suicide & Crisis Lifeline, and OpenAI said users outside the US would be pointed to local equivalents through directories such as Find A Helpline. OpenAI said it does not share ChatGPT conversations with the crisis lines, and that counsellors cannot see a user’s OpenAI account.

The feature arrived amid a cluster of wrongful-death and product-liability lawsuits against OpenAI, including the suit brought by the family of Adam Raine, which alleged ChatGPT had contributed to their son’s suicide, and followed OpenAI’s own September 2025 announcement of broader changes to how the model handles mental-health crises. It was one of several concrete product changes the company rolled out that autumn as it faced mounting legal and regulatory scrutiny of how conversational AI systems respond to users expressing suicidal ideation or severe distress, particularly minors, a set of concerns it addressed at greater length in a public statement on its litigation approach five days later.