Community safety report and new protections for teens

OpenAI is announcing two new initiatives to advance community safety: a Community Safeguards Report and additional restrictions on firearms content for minors. The company reports making over 200 law enforcement referrals in Q3 2026 for imminent threats of violence, relying on mental health and behavioral experts to assess difficult cases. The Community Safety Report, first to be published in early 2027 covering the second half of 2026 with updates every six months, will detail how OpenAI trains its systems to detect, investigate, and respond to safety issues, share anonymized examples, and highlight new mitigations. New policy boundaries restrict ChatGPT from providing detailed reconstructions of past mass attacks, including weapon details and comparisons that could amplify an attacker's notoriety, regardless of user age. ChatGPT will respond more conservatively when recent chats show warning signs of possible violence, and when conversations suggest fixation on violence, ChatGPT can automatically end chats and may temporarily pause accounts. For teens, further limits on weapons-related information are being added, and the strengthened age prediction system continually assesses signals to place users under 18 in the teen experience, with a privacy-protective verification process for adults. Parental controls now include alerts if a teen's account is deactivated for violence policy violations, with notifications for repeated violence discussions being developed. The safeguards will begin rolling out in the next few weeks.

(Source：OpenAI)