OpenAI Commits to a Community Safety Report From Early 2027, Updated Every Six Months, and Tightens ChatGPT Limits on Past Mass Attacks and Teen Weapons Questions

On October 30, 2026, OpenAI made two announcements about community safety: a new safety report it commits to publish, and tighter ChatGPT limits on violence-related information, with extra limits for teens. The post says OpenAI made over 200 law enforcement referrals in Q3 2026 and describes them as helping to disrupt violent plots; the count is OpenAI's own and the post gives no breakdown. The first report is due in early 2027, and the new safeguards are to begin rolling out in the next few weeks.

The report comes first. OpenAI commits to a Community Safety Report on violence prevention and other community safety issues, and on how well its safeguards work. It says the report will explain how its systems are trained to detect, investigate and respond to such issues, include anonymized examples of how violence-related harm shows up, highlight new mitigations and policy changes, and include relevant data. The first edition will cover the second half of 2026 and is due in early 2027, with updates every six months after that. The post uses the name Community Safety Report in its heading and Community Safeguards Report once in its opening, so the name is not fully settled. Nothing in the report can be checked yet, because it does not exist, and the post gives no effectiveness figure.

For all users, OpenAI is adding limits on past mass attacks and the people behind them. It says ChatGPT will give less detailed information in those areas even when the details are available from other online sources. The tightened policy restricts reconstructions of past mass attacks, including details about the weapons and equipment used and how they were obtained, and detailed comparisons of the scale of harm that could amplify an attacker's notoriety. People can still ask ChatGPT to help them understand these events, how victims and communities were affected, and how to recognize warning signs and prevent violence. The limits apply regardless of a user's age or reason for asking.

Two further changes concern warning signs. Building on a feature shared in May, in which ChatGPT can use short factual summaries of earlier warning signs, OpenAI says that when recent chats show signs of possible violence, ChatGPT will respond more conservatively with information that could be misused. It is also working to apply those stricter limits when it predicts that a user is under 18, even if the chats show no violent intent. When a conversation shows a person fixating on violence or on seriously harming someone, OpenAI says ChatGPT will be able to end the chat automatically, and if it ends several chats for that reason, the account may be paused temporarily. The post calls this an ongoing effort and says OpenAI is working with mental health and violence prevention experts on earlier, automated ways to intervene.

For teens, OpenAI adds limits on weapons-related information even when a question looks harmless. It extends existing rules to practical information about weapons and security systems that could be misused to facilitate violence, even with no sign of violent intent. Teens can still use ChatGPT to help prevent and de-escalate violence, to seek support and to reach a trusted adult. OpenAI says it strengthened its age prediction system, and that anyone wrongly placed in the under-18 experience can confirm adulthood through its partner Persona. Parental controls already include safety alerts, which may be sent when specially trained reviewers identify a serious self-harm concern or behavior that may be linked to disordered eating; in July, OpenAI extended the alerts to cases where a teen's account is deactivated for violating the violence policy. It says it is working on notifications for parents when it pauses a teen's account for repeatedly discussing violence.

These are announcements and commitments, and the measures roll out over the next few weeks. The post gives no numbers on how often the new limits trigger, how accurate age prediction is, or how many chats would be ended. For readers, the useful point is the concrete scope: past mass attacks, weapons questions from teens and repeated fixation on violence. The first real test will be the report due in early 2027.