OpenAI announced on October 30, 2026, that ChatGPT will limit details about past mass attacks for all users and practical weapons information for minors. The measures are set to begin rolling out over the following few weeks.
Fewer details about past attacks
The tighter rules will restrict reconstructions of past mass attacks regardless of a user’s age or reason for asking. ChatGPT is to provide less detail about the weapons and equipment used and how attackers obtained them. The restrictions also cover detailed comparisons of the scale of harm that could increase an attacker’s notoriety.
People will still be able to ask about the events, including to understand their effects on victims and communities. ChatGPT is also intended to continue helping users recognize warning signs and prevent violence.
Stricter responses for minors
For users under 18, OpenAI is expanding restrictions on practical information about weapons and security systems that could facilitate violence. The stricter rules are intended to apply even when a question shows no sign of violent intent. Minors will still be able to use ChatGPT to prevent or de-escalate violent situations, seek support and connect with a trusted adult.
The company is working to apply the stricter limits to people its system estimates are under 18, even if their conversations show no signs of violent intent. The age-prediction system now continually assesses new signals as people use ChatGPT and moves minors into the under-18 experience when appropriate. Adults placed there incorrectly can verify their age through OpenAI’s partner Persona.
Responding to warning signs
OpenAI is also developing its use of short, factual summaries of earlier warning signs. If recent conversations suggest possible violence, ChatGPT is intended to respond more cautiously when providing information that could be misused.
The company is working with mental health and violence prevention experts on an automated intervention. ChatGPT could end a conversation showing signs of fixation on violence or seriously harming another person; if it repeatedly ends chats for that reason, the company could temporarily pause the account.
OpenAI is preparing notifications for linked parental accounts if a minor’s account is paused because of repeated discussions of violence. Since July, it has notified parents when a child’s account is deactivated for violating rules on violence.
Report on safety measures
OpenAI has also promised a Community Safety Report covering violence prevention, other community safety issues and the effectiveness of its safeguards. The first report is due in early 2027 and will cover the second half of 2026, with updates planned every six months.
The report is intended to explain how the company trains systems to detect, investigate and address safety issues. It will include anonymized examples, new interventions, policy changes and related data. OpenAI said it made more than 200 referrals to law enforcement in the third quarter of 2026 as part of its work to alert authorities to imminent threats of violence.
