OpenAI Forms Another Safety Committee After Dismantling Prior Team

OpenAI has formed a safety and security committee led by company directors and the CEO to evaluate and recommend safeguards for the next 90 days as the company trains its next LLM; the committee will report recommendations to the board before public release. The announcement follows the resignation of a former safety executive and the disbanding and reassignment of a superalignment team, prompting experts to question whether the change adequately addresses broader AI risks such as accuracy, fairness, and transparency.