Google, Microsoft and xAI Agree to Test AI Models with US Commerce Department

New artificial intelligence tools and capabilities from Google, Microsoft and xAI will undergo testing by the US Department of Commerce before public release.

The tech firms agreed to voluntarily submit their models through Commerce's Center for AI Standards and Innovation, known as CAISI.

These pacts build on earlier agreements reached by companies including OpenAI and Anthropic during the Biden administration. AI models from all the companies will face evaluation for capabilities and security.

"These expanded industry collaborations help us scale our work in the public interest at a critical moment," said CAISI director Chris Fall.

The evaluations will cover testing, collaborative research and best practice development related to commercial AI systems.

Google's best-known AI tool is Gemini, developed through its DeepMind subsidiary. The chatbot appears widely on Google products and now serves US defense and military agencies.

Microsoft's best-known AI tool is CoPilot. xAI's sole AI product is Grok, a chatbot that drew widespread scrutiny for generating images that undressed people.

On Tuesday, CAISI reported conducting 40 prior evaluations of AI tools, including tests of certain state-of-the-art models that remain unreleased. The center did not identify any models blocked from public release.

Representatives from Google, Microsoft and SpaceX, the Elon Musk company that controls xAI, did not respond to requests for comment.

Expanding company participation in AI research and safety testing marks a shift for the Trump White House, which has largely avoided oversight or regulation of AI and technology firms.

Last year, President Donald Trump signed executive orders forming the basis of his administration's "AI Action Plan." He said the plan would remove red tape and onerous regulation around AI development to ensure the US wins through advancements and control of the technology.

The US military continues to expand its use of AI. Recent claims by Anthropic state it developed a model called Mythos that is too powerful for public release.

Senior Trump staff members met last month with Anthropic CEO Dario Amodei, as the BBC previously reported. The company faces a lawsuit from the US Department of Defense over its refusal to remove safety guardrails for government use of its models.