Anthropic CEO Dario Amodei confirmed late Wednesday that the company will legally challenge the U.S. Department of Defense’s decision to designate the artificial intelligence (AI) company as a “supply chain risk.”
The designation, delivered via a formal letter from Defense Secretary Pete Hegseth on Wednesday, follows a high-stakes standoff between the startup and the Pentagon. Central to the conflict is Anthropic’s refusal to modify its safety protocols to allow its Claude AI models to be used for fully autonomous weaponry and the mass surveillance of U.S. citizens.
“We do not believe this action is legally sound,” Amodei said, asserting that his company has “no choice” but to seek a judicial remedy. Despite the gravity of the label, Amodei emphasized that the designation’s scope is legally narrow.
According to Anthropic, the restriction applies specifically to contracts where Claude is used as a direct component of Department of Defense operations. It does not, the company argues, prohibit the use of Claude by private-sector customers who happen to hold separate defense contracts.
“The statute exists to protect the government rather than to punish a supplier,” Amodei said. Anthropic has pledged to assist with a smooth transition for the DoD over the next six months, but the upcoming court case will likely define the boundaries between AI safety ethics and national security requirements for years to come.
Ripples of the Pentagon’s decision were felt immediately across Silicon Valley. Microsoft Corp., a major partner and investor in Anthropic, said Thursday it will continue to offer Claude to its commercial and governmental clients, with the sole exception of the Defense Department.
“Our lawyers have concluded that Anthropic products… can remain available to our customers through platforms such as M365, GitHub, and Microsoft’s AI Foundry,” a Microsoft spokesperson told CNBC.
Such support is a critical lifeline for Anthropic. While some defense contractors have reportedly instructed employees to migrate away from Claude, Microsoft’s decision to keep the model integrated into Microsoft 365 Copilot and GitHub Copilot ensures Anthropic maintains its massive footprint in the enterprise and software development markets.
The political friction hasn’t slowed Anthropic’s commercial momentum. In fact, it may have fueled it. Following President Trump’s recent calls for federal agencies to drop the service, public interest surged. On Monday, Anthropic’s apps briefly went down due to “unprecedented demand” after hitting the top of the Apple App Store charts, surpassing rival OpenAI’s ChatGPT for the first time.
Bloomberg reports that Anthropic is on track for $20 billion in annual revenue, doubling its run rate from late 2025.
The timing of the designation is sensitive. Reports have confirmed that Anthropic models played a role in recent U.S. airstrikes in the Middle East. However, as talks between the Pentagon and Anthropic over domestic surveillance stalled, OpenAI reportedly stepped in, agreeing to run its own models for classified workloads.
