OpenAI plans community safety reports and tightens ChatGPT teen protections

Community safety report and new protections for teens

OpenAI announced a community safety reporting program and tighter limits on violent content in ChatGPT, including extra protections for teens, rolling out over the coming weeks. New limits for all users cut details on weapons, gear and acquisition methods used in past mass attacks; teens face stricter limits on weapons and practical safety-system information. The first community safety report, covering the second half of 2026, arrives in early 2027 and updates every six months. OpenAI is also building interventions that end high-risk conversations and pause accounts that repeatedly discuss violence.

Why it matters: OpenAI extends teen weapons limits to questions with no sign of violent intent, covering practical information that could be used to carry out violence.