Trump allies move to sideline AI safety advocates in new US oversight push

This piece captures a critical shift in how AI safety debates are being institutionalised in Washington. For several years, Bay Area nonprofits rooted in effective altruism effectively set the vocabulary around catastrophic AI risk. Now Trump aligned figures are trying to keep some of the guardrails, like third party audits, while cutting those same groups out of the room in favour of blue chip consultancies and industry friendly coalitions. That is not just a culture war story, it is a fight over who gets to define “safe enough” for frontier systems.

For the AGI race, losing influence for technically sophisticated but politically awkward safety labs could tilt the balance toward faster deployment with more performative oversight. Accenture or EY can build real audit capacity, but they are unlikely to push as hard for dramatic slowdowns or red lines as METR or FAR.AI. At the same time, companies like Anthropic and OpenAI are trying to maintain legitimacy with regulators while managing billion dollar investors and a White House that openly dismisses “woke AI”. The likely outcome is not a clean win for either side but a patchwork in which safety is partly co opted into standard consulting workstreams, even as more radical voices get pushed to the margins.