Multiple reports say the United States is urging Meta to agree to a framework for reviewing its artificial intelligence systems, as Washington focuses on national security risks posed by powerful AI models. According to reporting cited by outlets, the discussions reflect growing concern among U.S. officials about how advanced AI capabilities could be used or misused, including potential impacts on cybersecurity, information security, and other national security interests. The reports indicate that the push for AI reviews is part of broader efforts by the US government to increase oversight and risk management for leading AI developers. While the accounts do not detail specific technical requirements or the final scope of any review arrangement, they describe the effort as a move to establish expectations for scrutiny of AI systems before or as they are deployed. The reports also note that the concern is rising in Washington, with policymakers looking for ways to ensure that large AI providers take steps to mitigate security threats.