US Gov Presses Meta to Share AI Models for Review

What is the US government asking Meta to do?

The US government is pressing Meta to submit its AI models for evaluation by federal officials, according to The New York Times. The goal is to assess model capabilities and identify potential vulnerabilities. The government has been sending its requests to Meta via email.

Meta is currently the only major AI developer that has not voluntarily agreed to this review. Every other large AI company has already done so.

Which companies have already agreed to share their models?

Here's where each major AI developer stands on the federal review process:

Reuters reported that OpenAI and Anthropic are already working directly with the government to test unreleased models. Google, xAI, and Microsoft have agreed to provide early access to their new models to the Center for AI Standards and Innovation.

What is the Center for AI Standards and Innovation?

The Center for AI Standards and Innovation is a federal agency created by the Biden administration and staffed with technical experts to vet AI technology. It is now headed by Commerce Secretary Howard Lutnick. The center's role is to evaluate AI models before they are released to the public.

What has Meta said about the review request?

Meta spokesperson Francis Brennan issued a statement: "We share the [Trump] administration's goal of advancing US leadership on robust and secure frontier AI. While we are working through the details, we hope to sign the agreement soon."

You might also like

Agility Digit: Live at Amazon, GXO & Toyota

Nvidia Banned Chips Double in Price on China

CAISI Signs AI Reviews With Google, Microsoft

Meta Smart Glasses Switch to Muse Spark AI

The company has not yet formally agreed to the review process, but its public statement signals it does not intend to refuse outright.

What executive order created this review framework?

President Trump signed an executive order on June 2 to establish a framework for the federal government to evaluate AI releases. Under the order, the government has until the end of July to develop a formal review process. The target is for companies to give authorities up to 30 days to evaluate their AI technologies before making them public.

As we read the timeline, the formal process doesn't exist yet — but most major companies have been sharing models with the government voluntarily for months already, without waiting for the official framework.

What AI model has Meta released recently?

Meta launched its latest model, Muse Spark, in April. Muse Spark has two modes: "Instant" and "Thinking." The Thinking mode enables reasoning capabilities, giving the model extra time to work through a prompt for a more thorough answer. Engadget noted it is not quite as powerful as other companies' frontier models.

How does this fit into broader AI security concerns?

The federal government has been tightening its oversight of the AI industry. In mid-June, it ordered Anthropic to suspend access to its Mythos 5 and Fable 5 models for all foreign nationals, citing national security concerns. Anthropic blocked everyone's access to both models to comply with the directive.

Mythos is Anthropic's cybersecurity-focused AI model, available only to its Project Glasswing partners. Mythos 5 is the latest version. Fable 5 was designed to bring many of Mythos' capabilities to the public — Anthropic said it outperformed any previous model the company had launched, though it is less powerful than Mythos.

These moves signal a pattern of growing federal scrutiny over AI model capabilities, particularly around security risks. Builders tracking this space may also want to watch how AI security concerns are shaping infrastructure decisions more broadly, or how AI content integrity is becoming a policy flashpoint alongside model safety.

The pressure on Meta also comes as major players race to build and deploy more capable systems — a race that has drawn attention to AI hardware investments and the infrastructure behind frontier models.

Frequently asked questions

Why is the US government asking Meta to share its AI models? The government wants to assess the capabilities of Meta's AI models and identify potential security vulnerabilities. According to The New York Times, Meta is the only major AI developer that has not voluntarily submitted its models for federal review. The government has been making its requests via email, and a formal review process is expected to be completed by the end of July 2026.

What is the deadline for the federal AI review process? President Trump signed an executive order on June 2 establishing a framework for federal AI evaluation. Under that order, the government has until the end of July 2026 to develop a formal review process. The goal is for companies to give authorities up to 30 days to evaluate AI technologies before they are released to the public.

Has Meta refused to comply with the government's AI review request? Meta has not refused. Spokesperson Francis Brennan said the company shares the Trump administration's goal of advancing US AI leadership and that Meta "hopes to sign the agreement soon." The company says it is still working through the details of the arrangement.

What happened to Anthropic's Mythos 5 and Fable 5 models? In mid-June 2026, the US government ordered Anthropic to suspend access to its Mythos 5 and Fable 5 models for all foreign nationals, citing national security concerns. Anthropic blocked everyone's access to both models to ensure compliance. Mythos is a cybersecurity-focused AI available only to Project Glasswing partners; Fable 5 was a public-facing version of similar capabilities.

Which AI companies are already working with the US government on model reviews? OpenAI and Anthropic are working with the government to test unreleased models. Google, xAI, and Microsoft have agreed to give the Center for AI Standards and Innovation early access to their new models. Meta is the only major AI developer that has not yet formally agreed to participate in the review program, according to reporting from The New York Times and Reuters.