Pentagon's Red Flag: Anthropic a Supply Chain Risk

The Department of Defense has officially informed Anthropic's leadership that the company and its offerings have been classified as a supply chain risk, according to a Bloomberg report citing a senior department official.

This classification follows weeks of contention between the artificial intelligence research lab and the DoD. Anthropic CEO Dario Amodei has reportedly declined to permit the military to deploy its AI systems for widespread surveillance of U.S. citizens or to power fully autonomous weaponry lacking human involvement in targeting and firing decisions. The Department, conversely, has asserted that its application of AI should not be constrained by a private contractor.

Such supply chain risk designations are typically reserved for foreign adversaries. This specific label now mandates that any company or agency collaborating with the Pentagon must certify that it does not utilize Anthropic's AI models.

This determination by the Pentagon poses a significant threat to both Anthropic's operations and the Department's own strategic capabilities. Notably, Anthropic had been the sole frontier AI laboratory possessing systems deemed ready for classified applications. The U.S. military currently leverages Anthropic's Claude in its Iran campaign, where American forces employ AI tools for efficient data management in their operations. According to Bloomberg, Claude is a primary component integrated into Palantir’s Maven Smart System, which is critical for military operators in the Middle East.

Several critics have characterized the Department's action of labeling Anthropic a supply chain risk due to this dispute as an unprecedented step. Dean Ball, a former AI advisor to the Trump White House, controversially described the designation as a “death rattle” of the American republic. He contended that the government has forsaken strategic clarity and respect, opting instead for "thuggish" tribalism that disadvantages domestic innovators more severely than foreign adversaries.

In response, hundreds of employees from OpenAI and Google have publicly called upon the DoD to rescind its designation. They have also appealed to Congress to challenge what they view as an inappropriate exercise of authority against an American technology firm. Furthermore, these employees have urged their respective leaderships to unite in continuing to resist the DoD’s demands for the deployment of their AI models in domestic mass surveillance and for "autonomously killing people without human oversight."

TechCrunch has reportedly sought comment from Anthropic regarding these developments.

Amidst this ongoing dispute, OpenAI proceeded to finalize its own agreement with the Department, permitting the military to utilize its AI systems for “all lawful purposes.” This agreement has, however, prompted concern among some of OpenAI’s own employees, who worry that the ambiguous phrasing could inadvertently enable the very applications Anthropic sought to prevent.

Amodei has characterized the DoD’s actions as “retaliatory and punitive,” and he reportedly suggested that his refusal to publicly praise or contribute financially to President Trump played a role in escalating the dispute with the Pentagon. In contrast, OpenAI President Greg Brockman has been a vocal supporter of Trump, recently contributing $25 million to the MAGA Inc. Super PAC.