Recently, the US Department of Defense flagged Anthropic as a supply chain security risk, prompting federal agencies to phase out procurement of the company’s AI offerings. The move stems from the defense acquisition system’s security assessment procedures, not from Treasury Secretary Janet Yellen nor from sanctions targeting the financial system.
The DoD’s evaluation found that Anthropic retains functional modules in certain AI systems that could be leveraged for surveillance and autonomous weapons, failing to meet the mandatory controllability and transparency standards required in federal procurement. Consequently, the White House has ordered all federal agencies to suspend new contracts with Anthropic and to inventory existing systems that rely on the technology while developing replacement plans.
The impact extends beyond government departments to third-party contractors working with federal agencies. These firms must review their subcontracting agreements to ensure they do not inadvertently violate the federal procurement ban via indirect use of Anthropic technology, while also preparing data migration, system replacement, and continuity plans. Notably, this restriction applies solely within the federal procurement domain; private companies and NGOs may continue to use Anthropic products legally, although industries subject to strict compliance, such as finance and healthcare, may proactively adjust vendor strategies to mitigate risk.
During the technical phase-out process, agencies typically coordinate across their CIO, CISO, and privacy offices to conduct systematic risk assessments, evaluate alternatives, and perform security audits. The entire process must be fully documented, including technical equivalence evidence, vendor due diligence records, and transition logs, to support future audits.
Looking ahead, the ban could be lifted if Anthropic reaches a new compliance agreement on its security framework or if the White House and DoD reassess its risk profile. Any policy reversal would require formal notification and a restart of procurement compliance reviews. This move underscores another critical step by the US government in enforcing supply chain security over AI vendors and highlights the tightening regulatory environment for AI in defense and public safety.