OpenAI said on Friday that ChatGPT will be able to automatically end a chat when a conversation shows signs that someone is fixating on violence, and that it is placing further limits on weapons information for teens, in a post on its site. OpenAI said the safeguards it detailed “will begin rolling out in the next few weeks.” Bloomberg Law also reported the announcements.
OpenAI said that when a conversation shows signs “that someone is fixating on violence or the idea of seriously harming another person, ChatGPT will be able to automatically end the chat to interrupt that pattern in the moment.” If ChatGPT ends multiple chats for that reason, OpenAI said, it may temporarily pause the account. It said it is working with mental health and violence prevention experts on additional automated ways to intervene earlier, and called this an ongoing effort.
For users under 18, OpenAI said it is placing further limits on weapons-related information, even when a question appears harmless. It said it is extending existing restrictions to cover practical information about weapons and security systems that could be misused to facilitate violence, even when there is no sign of violent intent. Teens can still use ChatGPT to help prevent and de-escalate violence, seek support and connect with a trusted adult, OpenAI said.
OpenAI said it strengthened its age prediction system to keep assessing new signals and to move teens into the under-18 experience when appropriate. Users placed there incorrectly can confirm they are adults through Persona, its age verification partner. OpenAI said it is working to introduce notifications for parents if it pauses a teen’s account for repeatedly discussing violence; it did not say when. In July, it said, it expanded parental alerts to cover a teen’s account being deactivated for violating its violence policies.
For all users, regardless of age, OpenAI said it is restricting how ChatGPT responds to questions about past mass attacks, including details about weapons and equipment used and how they were obtained.
OpenAI said it brings imminent threats of violence to the attention of law enforcement and that it made over 200 law enforcement referrals in the third quarter of 2026. The figure is OpenAI’s own; the post does not break it down or say how it was verified.
OpenAI said it will publish its first Community Safety Report in early 2027, covering the second half of 2026, with updates every six months. It said the report will include anonymized examples, new mitigations and policy changes, and relevant data.
Related coverage: Texas Attorney General Ken Paxton’s lawsuit against OpenAI and Sam Altman over ChatGPT safety claims, filed Oct. 23, which OpenAI’s post does not mention, and OpenAI’s updated Model Spec.
Sources
- OpenAI: Community safety report and new protections for teens (October 30, 2026)
- Bloomberg Law (Sebastian Tong, Bloomberg News): OpenAI Expands ChatGPT Protection for Teens (Oct. 30, 2026, 4:22 PM UTC; headline and summary bullets only, paywalled)
- AI Tech Daily: Texas sues OpenAI and Sam Altman over ChatGPT safety claims
- AI Tech Daily: OpenAI says updated Model Spec reorganizes its guidance around five values and retires most detailed style guidance