OpenAI has endorsed a key part of a bipartisan House bill on AI safety. The FRONTIER Act comes from Reps. Jay Obernolte and Lori Trahan. It would create the first federal framework for AI safety in the United States. Large AI developers would have to let outside evaluators inside their companies to check models are built safely. Those reviewers are called independent verification organizations. The requirement would apply only to firms above high revenue and computing-spend thresholds. Chris Lehane, OpenAI's chief global affairs officer, said the company can support that provision. It is the first time OpenAI has backed such a federal mandate. Last week it also endorsed a California accreditation law for the same kind of reviewer.