The Trump administration is pressing Meta Platforms to voluntarily submit its artificial intelligence models for federal evaluation as national security officials escalate scrutiny over the technology sector. U.S. officials made the request through confidential emails with the social media giant, seeking access to its systems to evaluate their capabilities and identify potential technical vulnerabilities, according to an exclusive report published on Tuesday by The New York Times. The outreach marks the administration's latest effort to police advanced digital systems before they are released widely, ensuring they do not present risks to critical infrastructure or domestic security.
Meta Remains Sole Major U.S. AI Developer Without Federal Review Agreement
Citing four people familiar with the confidential request, The New York Times reported that Meta is the only primary U.S. developer of advanced artificial intelligence that has not reached an agreement to voluntarily share its models with the federal government. Other industry frontrunners—including OpenAI, Anthropic, Google, xAI, and Microsoft—have all established agreements to submit their technology for review by a federal oversight group known as the Center for AI Standards and Innovation. The government uses these reviews to test models for high-level safety risks, such as the potential to exploit critical software or aid in cyberattacks, prior to commercial rollout.
"We share the administration's goal of advancing U.S. leadership on robust and secure frontier A.I.," Francis Brennan, a Meta spokesman, said in a statement on Tuesday. "While we are working through the details, we hope to sign the agreement soon."
Trump Executive Order Mandates 30-Day Federal AI Review Period
The pressure on the parent company of Facebook and Instagram comes amid a broader regulatory clampdown on the fast-evolving artificial intelligence industry. Earlier this month, President Donald Trump signed an executive order granting the federal government responsibility for reviewing advanced AI systems. The order mandates tech companies to allow up to 30 days for federal officials to evaluate cutting-edge AI models before they are publicly distributed. The administration has set a deadline for the end of July to establish its formal review process.
Anthropic Ordered to Restrict Newest AI Model Access
Federal anxieties over the pace of AI development were highlighted less than two weeks ago when the government ordered tech developer Anthropic to restrict access to its newest model, citing immediate national security concerns. The directive to Anthropic represents the administration's willingness to enforce restrictions on AI releases when security risks are identified.
FAQ
What did the Trump administration request from Meta regarding AI models?
U.S. officials contacted Meta through confidential emails seeking access to its artificial intelligence systems to evaluate their capabilities and identify potential technical vulnerabilities, according to a report published on Tuesday by The New York Times.
Which major U.S. AI developers have federal review agreements?
OpenAI, Anthropic, Google, xAI, and Microsoft have all established agreements to submit their technology for review by the Center for AI Standards and Innovation. Meta is the only primary U.S. developer of advanced artificial intelligence that has not reached such an agreement.
What does the Trump executive order require for AI model releases?
The executive order signed earlier this month mandates tech companies to allow up to 30 days for federal officials to evaluate cutting-edge AI models before they are publicly distributed, with a deadline for the end of July to establish the formal review process.
