The Hidden Gatekeepers of AI Safety

The current oversight landscape is fractured and ad hoc. While an executive order signed last month promises a formal roadmap, key figures like former White House advisor Sriram Krishnan have explicitly ruled out the creation of an AI-focused FDA. Authority currently rests loosely with the Department of Commerce’s Center for AI Standards and Innovation, yet six cabinet agencies are still scrambling to define a final process by August.

Critics argue that this regulatory vacuum is filled by personal influence rather than rigorous policy. OpenAI CEO Sam Altman and President Greg Brockman have maintained close ties to the Trump administration, leading to public speculation that financial support and political lobbying have secured the company a lighter-touch approach. This contrasts sharply with the experience of Anthropic, whose model Fable faced a temporary public access ban following friction with the administration.

Industry veterans like Andy Konwinski warn that relying on opaque, private negotiations creates an existential risk. Without a standardized, transparent framework—or the inclusion of independent safety and alignment researchers—the industry risks losing public trust. As the sector continues to prioritize rapid commercialization to recoup massive training costs, the absence of an "open commons" for evaluation leaves the scientific community and the general public on the outside looking in.