Anthropic stands firm as Pentagon intensifies AI contract dispute

Anthropic has until Friday evening to either grant the U.S. military unrestricted access to its AI model or face escalating repercussions, according to Axios.

Defence Secretary Pete Hegseth told Anthropic CEO Dario Amodei during a meeting Tuesday morning that the Pentagon could either label Anthropic a “supply chain risk” — a designation more commonly applied to foreign adversaries — or turn to the Defence Production Act (DPA) to compel the company to build a version of its model that meets military requirements.

The Defence Production Act authorises the president to require companies to prioritise or expand production in support of national defence. In recent history, it was invoked during the COVID-19 pandemic to push companies such as General Motors and 3M to produce ventilators and masks.

Anthropic has repeatedly said it does not want its technology used for mass surveillance of Americans or for fully autonomous weapons, and it is refusing to walk back those positions.

Pentagon officials, for their part, have argued that the military’s use of technology should be governed by U.S. law and constitutional constraints, not by the internal use policies set by private vendors.

Using the DPA to fight over AI guardrails would represent a notable expansion of how the law is being applied in the modern era. It would also reflect a widening pattern of executive-branch volatility that has intensified in recent years, said Dean Ball, a senior fellow at the Foundation for American Innovation and a former senior AI policy advisor in Trump’s White House.

“It would basically be the government saying, ‘If you disagree with us politically, we’re going to try to put you out of business,’” Ball said.

The standoff is playing out amid broader ideological tension, with some administration figures — including AI czar David Sacks — publicly attacking Anthropic’s safety policies as “woke.”

“Any reasonable, responsible investor or corporate manager is going to look at this and think the U.S. is no longer a stable place to do business,” Ball said. “This is attacking the very core of what makes America such an important hub of global commerce. We’ve always had a stable and predictable legal system.”

The dispute has turned into a high-stakes game of chicken, and Anthropic may not be the side that backs down. Reuters reported that the company does not intend to loosen its usage restrictions.

Several reports have said Anthropic is the only frontier AI lab with classified access to the Department of Defence. The Pentagon does not currently have a clear replacement lined up. However, it has reportedly reached an agreement to use xAI’s Grok in classified environments.

That lack of redundancy helps explain the Pentagon’s hardline posture, Ball argued.

“If Anthropiccancelledd the contract tomorrow, it would be a serious problem for the DOD,” he said, adding that the agency appears to be falling short of a National Security Memorandum from the late Biden administration that directed federal agencies to avoid dependence on a single classified-ready frontier AI system.

“The DOD has no backups. This is a single-vendor situation here,” he continued. “They can’t fix that overnight.”