OpenAI said new safeguards will start rolling out in the next few weeks, adding stricter limits on violence and weapons-related answers for all users and especially for teens.
Under the changes, ChatGPT will give less detail on past mass attacks, including weapon acquisition, attack reconstructions and comparisons that could amplify an attacker’s notoriety.
For users flagged by recent chats or predicted to be under 18, the system will respond more conservatively; repeated violence-focused conversations can trigger automatic chat termination and temporary account pauses.
OpenAI also said it made more than 200 law-enforcement referrals in Q3 2026 and will publish its first Community Safety Report in early 2027, covering the second half of 2026.