Microsoft CEO Satya Nadella speaks at the Microsoft AI Tour event in Munich, Germany on February 25, 2026.
Sven Hoppe | Picture Alliance | Getty Images
microsoft announced Thursday that it will continue to incorporate startup Anthropic’s artificial intelligence technology into its products for customers other than the U.S. Department of the Army.
Earlier Thursday, federal regulators notified Anthropic that they were designating the company as a supply chain risk. Anthropic later said it intended to challenge the move in court.
Last week, President Donald Trump asked federal agencies to stop using Anthropic, and Army Secretary Pete Hegseth said Anthropic could continue to serve the Department of Defense for up to six months. CNBC has confirmed that human models played a role in US airstrikes against Iran in recent days.
Shortly after negotiations between the Army Corps and Anthropic on domestic mass surveillance and fully autonomous weapons broke down on Friday, Anthropic’s rival OpenAI announced that the Pentagon had agreed to run its models against classified workloads.
Microsoft is the first major company to say it will continue working with Anthropic after the Pentagon’s action. Some defense technology companies have told their employees to stop using Anthropic’s Claude model and move to alternative models.
“Our lawyers have reviewed this designation and have concluded that Anthropic products, including Claude, will continue to be available to our non-Department of the Army customers through platforms such as M365, GitHub, and Microsoft’s AI Foundry, and that they can continue to work with Anthropic on non-defense-related projects,” a Microsoft spokesperson told CNBC in an email.
Microsoft provides technology to various US government agencies. Microsoft 365 productivity software is widely used within the Department of the Army. Microsoft announced in September that it would integrate Anthropic’s generative artificial intelligence models, along with models from OpenAI, into the Microsoft 365 Copilot add-on for Microsoft 365 subscriptions.
Many software engineers rely on Anthropic’s Claude models to write their source code, and they are available on GitHub Copilot AI software development services, along with OpenAI’s competing Codex models.
Microsoft CEO Satya Nadella initially invested in and sought to work closely with OpenAI. More recently, Nadella has made AI models created by other developers available to Microsoft products.
“Choose your model!” In his October X post, Nadella introduced the ability to switch between Anthropic and OpenAI models in Microsoft 365 Copilot.
Microsoft announced in November that Anthropic had committed $30 billion to Microsoft’s Azure cloud services, while Microsoft had agreed to invest up to $5 billion in Anthropic. The numbers for Microsoft’s OpenAI partnership are even higher. As of October, Microsoft held a $135 billion stake in OpenAI’s public interest corporation, and OpenAI had committed $250 billion to Azure.
—CNBC’s Ashley Capoot contributed to this report.
