Trump Admin Pushes Meta to Submit AI Models for Safety Review
Meta remains the only major US AI developer without a voluntary government safety review agreement.
The Trump administration is leaning on Meta to voluntarily submit its AI models for government safety evaluations. According to the New York Times, Meta stands alone as the only major American AI developer that hasn't signed on to the review process.
Federal officials are ramping up pressure on the holdout just weeks after ordering Anthropic to pull its latest model — a move that signaled the administration isn't afraid to flex its muscle on AI oversight.
The voluntary review framework has already secured buy-in from every other significant US AI player. Meta's resistance puts the company in an increasingly awkward position as Washington tightens its grip on the AI sector.
Whether Meta caves or continues to dig in could set a major precedent for how open-source AI development intersects with government safety mandates going forward.