Microsoft Opens Its AI to the U.S. Government — Before Anyone Else Gets to See It

Microsoft has agreed to give the U.S. government early access to its AI models — before they are released to the public.

The arrangement, formalised with the Commerce Department's Center for AI Standards and Innovation, allows government agencies to assess capabilities and identify security vulnerabilities ahead of commercial launch.
Google made the same agreement on the same day. Together, they join OpenAI and Anthropic in what is rapidly becoming a formalised government preview programme for frontier AI — part of President Trump's AI Action Plan.
For Microsoft, this is both a compliance move and a strategic one. Azure remains the infrastructure backbone for a significant portion of OpenAI's operations.

Microsoft Copilot is now embedded across the entire Microsoft 365 suite — used by hundreds of millions of enterprise users worldwide.

The company spent $31.9 billion on AI infrastructure in Q1 2026 alone.
But investors are asking harder questions. Azure growth is strong, but is Microsoft translating its AI spending into enough product leadership? The comparison with Google is unflattering: Google Cloud grew 63% to $20 billion. Azure's numbers, while solid, have not triggered the same market enthusiasm.
Meanwhile, Microsoft offered voluntary retirement to 8,750 U.S. employees — about 7% of its domestic workforce — as its finance chief signalled headcount would decline this year in the name of "pace and agility."

The European perspective: Microsoft's agreement to give Washington early access to AI models raises a pointed question for Brussels: will European regulators get the same access — or will the EU be the last to know what these systems can do?