Skip to content
Neural Daily

PolicyNews

ChatGPT tightens violence safeguards and teen protections

🇭🇺 Olvasd el magyarul →

OpenAI is tightening how ChatGPT handles violence-related requests: the model can now end a conversation on its own in some cases, and minors get less weapons-related information. Starting in early 2027, the company will publish a public community safety report every six months.

By: Mikola Máté Patrik · · 2 min read

OpenAI title card reading Community Safety Report and New Protections for Teens on a yellow-green gradient
Image: OpenAI · Press kit · Official social image from the OpenAI announcement · Source

What changes

OpenAI made two announcements on October 30 (OpenAI): it is launching a Community Safety Report and further restricting weapons-related content for minors. The company says it made over 200 law enforcement referrals for imminent threats of violence in Q3 2026.

  • ChatGPT will give less detailed answers about past mass attacks. It will not reconstruct an attack, the weapons and equipment used or how they were obtained, and it will not compare attacks by scale of harm. This applies regardless of a user's age or reason for asking.
  • When recent chats show warning signs of possible violence, the model responds more conservatively with information that could be misused. The same stricter limits apply when the system predicts a user is under 18.
  • When a conversation shows signs that someone is fixating on violence or on seriously harming another person, ChatGPT can automatically end the chat. After multiple such endings, the account may be temporarily paused.

Teens and parents

ChatGPT for Teens now restricts practical information about weapons and security systems even when there is no sign of violent intent. The age prediction system keeps assessing new signals and moves users into the under-18 experience when appropriate. Adults placed there by mistake can verify their age through Persona, OpenAI's verification partner.

Since July, parents with linked accounts get an alert if their teen's account is deactivated for violating the violence policies. OpenAI is working on notifications for when a teen's account is paused for repeatedly discussing violence.

The report

The first Community Safety Report is due in early 2027, covering the second half of 2026, with updates every six months after that. OpenAI promises anonymized examples, new mitigations, policy changes and data on how well its safeguards work.

Why it matters

The new safeguards roll out over the next few weeks. Anyone using ChatGPT in education, youth work or security research should expect terser answers on violence and weapons topics, and occasionally an ended chat. The six-monthly report should at least make it possible to measure how these protections work in practice.

    Produced with the help of AI, reviewed by an editor. How we make our articles

    Reviewed and edited by the Neural Daily desk. About

    Related articles