OpenAI adds a direct board route for AI safety concerns

OpenAI personnel can now raise AI safety concerns or AI safety policy violations directly with the OpenAI Foundation Board’s Safety and Security Committee. The [October 30 update](https://openai.com/index/openai‑raising‑concerns‑policy/) changes an existing policy published in January. Managers, HR, Compliance and Legal remain established channels. OpenAI says its policy prohibits retaliation and permits reports to government agencies. Direct committee escalation is an additional internal route, not a replacement for those options.

The [linked policy](https://cdn.openai.com/policies/openai‑raising‑concerns‑policy.pdf#page=7) permits direct SSC escalation or the Integrity Line. For safety concerns and policy violations sent through that line, Compliance escalates them to the committee. Both routes lead to safety oversight. SSC consideration of an AI safety concern is described as input to safety decisions, not a formal investigation. Ordinary misconduct has a separate investigation process. The policy also permits [outside‑authority reporting](https://cdn.openai.com/policies/openai‑raising‑concerns‑policy.pdf#page=6) without prior notice to OpenAI. That keeps external reporting separate from the company’s internal escalation paths. OpenAI’s San Francisco offices in June 2025. Archival company context, not a committee meeting.