OpenAI Announces Safety Report and Teen Protections

OpenAI announced that it is launching a Community Safeguards Report and adding restrictions to ChatGPT concerning firearms content for minors and detailed information about past mass attacks. The company said the updated policies apply across the platform, alongside specific safeguards directed at users under 18.

OpenAI reported that it made over 200 law enforcement referrals regarding imminent threats of violence in Q3 2026. The company plans to publish its first Community Safety Report in early 2027 covering the second half of 2026, followed by updates published every six months.

Under the updated policies, ChatGPT is restricted from providing detailed reconstructions of past mass attacks, regardless of user age or reason for asking. The model is directed to limit detailed information regarding weapons, equipment, how items were obtained, and comparisons on scale of harm that could amplify an attacker's notoriety. Users can still ask ChatGPT questions to help understand these events, how victims and communities were impacted, and how to recognize warning signs to prevent future violence.

When recent chats show warning signs of possible violence, ChatGPT will respond more conservatively when sharing information that could be misused. OpenAI said it is also working to apply these stricter limits when predicting a user is under 18, even without signs of violent intent. If a conversation shows signs of fixation on violence or seriously harming another person, ChatGPT will be able to automatically end the chat, and OpenAI may temporarily pause accounts with multiple terminated chats.

For teen users, OpenAI is extending protections to restrict practical information about weapons and security systems that could be misused to facilitate violence. The company's age prediction system continuously evaluates signals to move teen accounts into the under 18 experience, with age verification handled through Persona. OpenAI is also working to introduce notifications to parents if a teen's account is paused for repeatedly discussing violence, building on alert features added in July for account deactivations. According to OpenAI, these safeguards will begin rolling out in the next few weeks.