Pentagon Halts Use of Anthropic’s Claude AI After Earlier Supply‑Chain Blacklist

The U.S. Department of Defense announced this week that it will no longer employ AI tools from Anthropic, the startup behind the Claude language model, marking a reversal of practice just months after the agency placed the company on a supply‑chain risk list.

Defense officials first labeled Anthropic a "supply chain risk" earlier in the year, citing concerns that the company's technology could expose sensitive data or be vulnerable to foreign influence. The designation placed Anthropic among a group of commercial vendors whose products would require additional scrutiny before integration into military systems.

Nevertheless, several internal sources confirmed that the Claude model had become widely used across multiple Pentagon units in the months following the blacklist. The model was reportedly embedded in a range of applications, from intelligence analysis to logistics planning, because of its strong performance on natural‑language tasks and the speed with which it could be deployed.

The decision to pull the tools underscores the tension between the defense department’s need for cutting‑edge AI capabilities and its mandate to safeguard the security of its supply chain. While commercial AI offerings can accelerate mission‑critical workflows, they also raise questions about data provenance, model transparency, and potential backdoors, especially when the providers are relatively new and privately held.

Looking ahead, the Pentagon says it will conduct a comprehensive review of its AI procurement policies, emphasizing stricter vetting procedures and greater reliance on vetted, domestically controlled technologies. Lawmakers and oversight bodies are expected to monitor the process closely, as the department seeks to balance rapid innovation with the imperative to protect national security.