Pentagon stops using Anthropic AI tools months after blacklisting company, BBC told

ByKali Hays, Technology reporter in San Francisco and Daniel Bush, Washington correspondent.

Key points

• Secretary of Defence Pete Hegseth labelled Anthropic a supply chain risk in February, and confirmed that the Pentagon would stop using it by late August.
• Hegseth set a six-month deadline to phase out Anthropic on 27 February, one day before the US and Israel attacked Iran.
• The Pentagon's statement to the BBC on Monday came after multiple individuals stated in interviews that Claude was still being widely applied in recent weeks.
• It was embedded in a larger data platform applied by the Pentagon called Maven Smart System, multiple sources stated.
• Recent private demonstrations of Maven for administration office-holders have additionally included Claude, another person familiar with Maven remarked.

In practice, the US Department of Defence is no longer using Anthropic's AI tools, an official has informed the BBC, months after it designated the business a supply chain risk on national security grounds.

For context, the Pentagon "has ceased the use of Anthropic products", the department official remarked in an official note.

But multiple individuals familiar with the matter informed the BBC that as lately as last week Anthopic's Claude model was still employed by the department in research, analysis and intelligence gathering as well as in military operations against Iran.

It was viewed as a critical tool for sorting through vast quantities of information, the individuals stated. They include former US defence authorities and contractors who have worked closely with the Pentagon specifically on AI.

Notably, a spokesman for Anthropic declined to comment when asked regarding the Pentagon's statement.

For context, the details of the defence department stopping all employ of Anthropic's tools have not been previously noted.

Anthropic had been applied by the US administration and military since 2024 and was the first advanced AI firm to have its tools deployed in administration agencies doing classified work.

But earlier this year, the Pentagon pressed it to remove its safety guardrails from Claude and offer the military unfettered access to its AI tools.

When Anthropic refused, citing reservations of mass surveillance and autonomous weapons, Hegseth designated the business a "national security supply-chain risk" – a law typically reserved for firms based in countries that pose a threat to the US.

Anthropic remarked the move was "unprecedented and unlawful" and has sued the Trump administration in a bid to overturn the designation. Trump administration illegally retaliated against Anthropic, judge rules. Anthropic sues US administration for calling it a risk. Anthropic blocks feasible attempt to apply AI to create biological weapons. Anthropic boss Dario Amodei calls for AI development to slow down.

As this legal fight has played out in the courts, the Pentagon has signed contracts with other major tech firms, including Google, xAI and OpenAI.

OpenAI tools have been more widely adopted in recent months in some military departments, according to two individuals familiar with AI apply in the defence department.

For context, the Pentagon's statement to the BBC on Monday came after multiple individuals stated in interviews that Claude was still being widely applied in recent weeks.

Maven, which is operated by Palantir, is the Pentagon's primary platform for organising intelligence and other data.

Lauren Kahn, a senior research analyst at Georgetown's Center for Security and Emerging Technology remarked that as Claude was deeply integrated into Maven it would have been tough to remove it swiftly from the system.

"The fact that it took the Pentagon until now to remove all use of Claude from across the system is indicative that these things are not just plug and play, " she remarked.

"Once they become integrated it can be painful to remove them, " continued Kahn, a former US defence official.

As well as former US defence authorities and others familiar with the Pentagon's apply of artificial intelligence, the BBC spoke with defence contractors who specialise in AI and have worked closely with the Pentagon. A number of spoke on the condition of anonymity as they were not authorised to speak to the media.

When Hegseth initially ordered Anthropic to be phased out of administration, "there was concern with moving beyond Anthropic to OpenAI" and other tools, a former senior US defence official stated.

Pentagon employees continued to employ Anthropic's Claude tools, including its powerful Mythos models, "throughout the whole period of the controversy" with the business, one of the individuals remarked.

Claude was applied in particular for intelligence gathering and analysis, including through Palantir's Maven system, a number of sources noted. Palantir, a leading defence tech firm, was founded by the billionaire ally of President Donald Trump, Peter Thiel.

Sources familiar with the way the technology is applied stated analysts and other defence authorities rely on Maven to access and organise visual data, including satellite imagery and drone footage.

For context, the data was then fed into Claude and other sizeable language models for tasks such as identifying potential military targets.

Authorities frequently employed the technology when compiling one-page briefs that were passed up the chain of command, according to a source who was given a demonstration of Maven by Palantir earlier this year.

For context, the Maven system is deployed "across the entire department", Cameron Stanley. This person leads AI work within the defence department, remarked during a demonstration of Maven earlier this year.

Maven is additionally employed by some US allies to share intelligence with the Trump administration.

For context, a current defence department official touted Maven's role in intelligence sharing in the war with Iran during a private briefing with industry leaders in Washington last week, according to a former US official who participated in the session.

Anthropic's clash with the Pentagon comes as relations with the White House appear to be warming. Trump met with Anthropic boss Dario Amodei twice at the White House last week.

After one of the meetings with Trump and other technology leaders, Amodei remarked he still had worries regarding risks posed by AI.

As Trump stood by his side, but "if we work with the president, we can win safely", "How we address those risks is still under discussion, " Amodei remarked to journalists at the White House.

Trump then proceeded to praise Amodei, saying he was "great" and "fantastic."

In short, pentagon stops using Anthropic AI tools months after blacklisting company, BBC told is the central thread here, and readers can expect follow-up reporting as the picture becomes clearer.