Tech Workers Press DOD, Congress to Drop Anthropic Label Over Supply‑Chain Risks

TechCrunch reported that a coalition of more than 200 engineers, data‑scientists and former defense contractors signed an open letter this week urging the Department of Defense to quietly rescind its “supply‑chain risk” designation on Anthropic. The signatories argue that the label, first applied in a classified risk‑assessment memo leaked to the press, threatens to isolate a company that has become a key provider of large‑language‑model services to federal agencies. “We’re not asking for a blanket exemption,” the letter reads, “just a reconsideration that reflects Anthropic’s track record of compliance and its growing role in government‑grade AI workloads.” The petition, which was circulated on a public GitHub repository, cites the company’s recent certifications under the DoD’s Cloud Computing Security Requirements Guide as evidence that the risk assessment was outdated.

The timing of the appeal dovetails with a broader push by the White House to engage “big‑tech” firms on energy‑cost concerns. Reuters noted that the administration will host a summit on March 4 that includes Microsoft, Meta and Anthropic alongside other data‑center operators. The meeting is intended to build on earlier commitments from Microsoft to power its AI clusters with renewable energy, a pledge that the White House hopes to extend across the sector. By placing Anthropic on the same agenda as the cloud giants, officials signal that the company’s infrastructure footprint is now viewed as strategically significant, even as the Pentagon’s internal memo continues to flag it as a potential supply‑chain vulnerability.

Anthropic’s inclusion in the upcoming White House gathering underscores a paradox: the firm is simultaneously being courted for its AI capabilities while being labeled a risk by the very agency that could be its biggest customer. Reuters’ coverage of Google’s AI‑driven transformation of its Cloud business highlights how quickly the market can shift when a vendor’s models become a growth engine for a tech conglomerate. The same dynamics are at play in Washington, where the Department of Defense is racing to integrate generative AI into mission‑critical systems. Critics of the risk label argue that it could force the DoD to turn to less vetted, open‑source alternatives, thereby eroding the security guarantees that a vetted commercial partner like Anthropic can provide.

The open letter also references the broader political climate surrounding defense‑tech procurement. A recent Reuters analysis of former President Trump’s “Musk‑led efficiency drive” warned that aggressive cost‑cutting could spur new partnerships between the Pentagon and private AI firms willing to build their own power plants or secure low‑cost electricity. By positioning Anthropic as a compliant, energy‑conscious partner, the signatories hope to align the company with the administration’s cost‑reduction agenda while defusing the supply‑chain narrative. They point out that Anthropic’s recent partnership with Microsoft to run its Claude models on Azure’s renewable‑energy‑backed clusters demonstrates a concrete step toward meeting both security and sustainability criteria.

If the Department of Defense heeds the workers’ plea, the immediate effect would be a quiet removal of the “supply‑chain risk” tag, allowing Anthropic to continue bidding on classified contracts without the stigma of a security flag. More importantly, it would set a precedent for how the federal government evaluates AI vendors: moving from blanket risk designations toward nuanced assessments that weigh compliance certifications, energy footprints and operational transparency. As the open letter’s authors conclude, “The real risk is not the technology itself, but the loss of a trusted partner in a landscape that increasingly demands secure, scalable AI.”