The US Department of Defence has ceased using Anthropic's AI tools, an official confirmed to the BBC. This move comes months after the Pentagon designated Anthropic a supply chain risk on national security grounds. Secretary of Defence Pete Hegseth had announced in February that the Pentagon would stop using Anthropic's tools by late August. The reasons for the delay in implementation are unclear.
Anthropic's AI tools, including its Claude model, were used by the Pentagon for research, analysis, and intelligence gathering, as well as in military operations against Iran. The Claude model was viewed as a critical tool for sorting through vast quantities of information. Multiple people familiar with the matter, including former US defence officials and contractors, confirmed that Anthropic's tools were still being used by the department as recently as last week.
The Pentagon's decision to stop using Anthropic's tools was made after the company refused to remove safety guardrails from its Claude model, which the Pentagon wanted for unfettered access to its AI products. Anthropic had cited concerns over mass surveillance and autonomous weapons. In response, Hegseth designated Anthropic a national security supply-chain risk, a designation typically reserved for companies based in countries that pose a threat to the US.
Anthropic has sued the Trump administration to overturn the designation, calling it "unprecedented and unlawful". The company had been used by the US government and military since 2024 and was the first advanced AI company to have its tools deployed in government agencies doing classified work. The Pentagon has since signed contracts with other major tech companies, including Google, xAI, and OpenAI.
OpenAI tools have been more widely adopted in recent months in some military departments, according to people familiar with AI use in the defence department. The Pentagon's primary platform for organising intelligence and other data, Maven Smart System, operated by Palantir, had deeply integrated Anthropic's Claude model, making it difficult to remove quickly. Removing the tool from the system was a complex process, according to Lauren Kahn, a senior research analyst at Georgetown's Center for Security and Emerging Technology.
The use of Anthropic's tools was widespread, with analysts and defence officials relying on Maven to access and organise visual data, including satellite imagery and drone footage. The data was then fed into Claude and other large language models for tasks such as identifying potential military targets. Officials frequently used the technology when compiling one-page briefs that were passed up the chain of command.
Anthropic's clash with the Pentagon comes as relations with the White House appear to be warming. Anthropic boss Dario Amodei met with President Trump twice at the White House last week, expressing concerns about risks posed by AI. Amodei stated that working with the president could help mitigate these risks. Trump praised Amodei, calling him "great" and "fantastic".
Key points
- The Pentagon has ceased using Anthropic's AI tools due to national security risks.
- Anthropic's tools were used for research, analysis, and intelligence gathering, as well as in military operations against Iran.
- The company's clash with the Pentagon comes as relations with the White House appear to be warming.
