From the source
OpenAI is tightening ChatGPT's rules on violent content and adding teen protections, alongside a new Community Safeguards Report on violence prevention.
- The company says it made over 200 law enforcement referrals in Q3 2026, bringing imminent threats of violence to authorities while relying on mental health and behavioral experts to assess difficult cases.
- ChatGPT will now give less detail on past mass attacks and their perpetrators, restricting reconstructions of attacks, weapon and equipment details, and comparisons of scale of harm that could amplify an attacker's notoriety.
- Those limits apply regardless of a user's age or reason for asking, though people can still ask about how victims and communities were affected and about recognizing warning signs.
- When recent chats show warning signs of possible violence, ChatGPT will respond more conservatively, and stricter limits will also apply when OpenAI predicts a user is under 18 even without signs of violent intent.
- ChatGPT will be able to automatically end a chat when a conversation suggests someone is fixating on violence, and repeated endings may lead to a temporary account pause.
- For teens, OpenAI is further limiting weapons-related information, including practical details about weapons and security systems that could be misused, while teens can still seek help preventing violence and connect with a trusted adult.
