OpenAI Restricts Teen Weapons Info, Auto-Ends Violence Chats

OpenAI said it made more than 200 referrals to law enforcement in the third quarter of 2026 involving potential threats of violence. The company disclosed the figure on October 30 as part of a broader violence-prevention push that will restrict weapons information for teenage ChatGPT users and automatically end conversations with users who appear to be fixating on violence, Reuters reported.

For users under 18, OpenAI said ChatGPT will now limit practical information about weapons and security systems "even when a question appears harmless," extending an existing ban on helping teens obtain weapons to cover queries with no sign of violent intent. The company said it has strengthened its age-prediction system to move users it estimates are under 18 into the teen experience, and it plans to notify parents if their teen's account is paused for repeatedly discussing violence. Teens can still use ChatGPT to help prevent and de-escalate violence, seek support, and connect with a trusted adult.

For all users, ChatGPT will be able to "automatically end chats where someone appears to be fixating on violence," and OpenAI may temporarily pause accounts after multiple such chats. The company is also "limiting detail on past mass attacks, including the weapons used and how they were obtained," directing the chatbot to provide less than it could, even where the material sits elsewhere online.

Alongside the product changes, OpenAI said it will publish a Community Safety Report on violence prevention every six months. The first edition is planned for early 2027 and will cover the second half of 2026, with anonymized examples of how violence-related harm shows up on the platform plus data on the effectiveness of its safeguards.

It is the company's second safety-focused announcement in two days, following a Wednesday rework of its Model Spec around five values. The topic has been crowded on our radar: we have tracked 392 OpenAI stories in the last 90 days and 351 on AI safety over the same window.