ChatGPT will end chats fixated on violence and limit weapons information for teens

OpenAI announced on October 30, 2026, a set of new restrictions on how ChatGPT handles violence. The chatbot will be able to end a conversation in which someone appears fixated on seriously harming another person. Teenagers will get less practical information about weapons, even when a question looks harmless. The company says the changes will begin rolling out in the next few weeks.

Limits that apply to every user

The broadest change concerns past mass attacks. ChatGPT will give less detail about these attacks and the people who carried them out, even when the information is available elsewhere online. OpenAI is restricting reconstructions of attacks, including 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. Explaining what happened, how victims and communities were affected, and how to recognize warning signs remains allowed.

Earlier conversations will also shape answers. OpenAI has already described how ChatGPT can use short, factual summaries of earlier warning signs when it responds in a later chat. Building on that, when recent chats show signs of possible violence, ChatGPT will be more conservative about sharing information that could be misused.

When a conversation shows that someone is fixating on violence or on seriously harming another person, ChatGPT will be able to end that chat automatically. If it ends several chats for that reason, OpenAI may temporarily pause the account. Accounts found to have violated the company's policies on violent activity are banned.

Stricter limits for users under 18

ChatGPT for Teens, the version of the service with extra safeguards for users under 18, will restrict practical information about weapons and security systems that could be misused to facilitate violence. The restriction applies even when nothing suggests violent intent. Teenagers can still use ChatGPT to prevent or de-escalate violence, seek support, and reach a trusted adult. OpenAI is also working to apply the stricter limits to anyone it predicts is under 18, even when their chats show no sign of violent intent.

Who counts as a teenager is decided by age prediction, a system that estimates age from signals such as behavior patterns. OpenAI says the system now keeps assessing new signals as people use ChatGPT and moves them into the under-18 experience when appropriate. Adults placed there by mistake can confirm their age through Persona, OpenAI's age verification partner, and have the extra protections removed.

Parents who link their account to a teenager's through parental controls already receive safety alerts. Since July 2026, they are also notified when the teenager's account is deactivated for violating the rules on violence. OpenAI is working on a further notice for when a teenager's account is paused for repeatedly discussing violence.

A public report from 2027

OpenAI has also committed to a Community Safety Report on violence prevention and how well its safeguards work. The first edition is due in early 2027 and will cover the second half of 2026, with updates every six months after that. The company says the report will include anonymized examples, new mitigations and policy changes, and supporting data.

OpenAI says it made more than 200 referrals to law enforcement over imminent threats of violence in the third quarter of 2026. The figure comes from the company and has not been independently verified.