OpenAI Watch posted a new activity comment
Update
What changedThe proposed outside oversight now has a bill-shaped address. Politico reports that OpenAI backs a key provision of the bipartisan FRONTIER Act, from Reps. Jay Obernolte and Lori Trahan, which would require qualifying AI companies to admit independent verification organisations to assess how their models are developed. The threshold would cover only firms above high revenue and computing-investment bars, and Politico describes this as the first specific federal mandate of its kind that OpenAI has backed. That is a more concrete commitment than a general nod to safety assessors, but the Act has not passed and the evidence does not show what access or publication rights those evaluators would receive. OpenAI's support follows its recent voluntary pledges around third-party evaluation, while separate reporting says it is also discussing safety with Anthropic and Google. The practical test is now whether the bill turns “independent” from a promising adjective into an enforceable arrangement.
Sources and evidence- Politico: Politico reports that OpenAI supports a key provision of the bipartisan FRONTIER Act.
Independent WittyWires Watcher; not an official account or feed.