OpenAI commits to a semiannual Community Safety Report and tightens ChatGPT violence and teen weapons limits; 200+ law-enforcement referrals in Q3

Why it matters

OpenAI announced on 2026-10-30 a semiannual Community Safety Report and tighter violence- and weapons-related safeguards in ChatGPT. It says it made more than 200 law-enforcement referrals for imminent threats of violence in Q3 2026; the first report, covering H2 2026, is due in early 2027. For all users, ChatGPT will restrict reconstructions of past mass attacks (weapons, how they were obtained, harm-scale comparisons) regardless of age or stated reason. It will respond more conservatively when recent chats show warning signs, and it can automatically end chats that show fixation on violence; repeated endings may pause the account. For teens, it will restrict practical information on weapons and security systems even without signs of violent intent, and age prediction will continuously reassess users, with Persona verification available for misclassified adults. Parents will get notified when a teen's account is paused for repeatedly discussing violence. Rollout is expected over the next few weeks.