US Demands Meta Accept Government AI Safety Tests

Federal officials are pushing Meta to agree to government safety reviews of its artificial intelligence models. The company is the only major tech firm that has not yet allowed such evaluations. The pressure comes weeks after regulators ordered Anthropic to remove its newest AI model from the market. That action marked a significant escalation in the government’s oversight of the fast-growing technology. Officials argue that independent testing is needed to ensure AI systems do not pose risks to users or national security. Meta has not publicly commented on the request.