The US Department of Defence is no longer using Anthropic's AI tools, an official has told the BBC, months after it designated the company a supply chain risk on national security grounds.
The Pentagon "has ceased the use of Anthropic products", the department official said in a statement.
Secretary of Defence Pete Hegseth labelled Anthropic a supply chain risk in February, and announced that the Pentagon would stop using it by late August.
But multiple people familiar with the matter told the BBC that as recently as last week Anthopic's Claude model was still used 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 people said. They include former US defence officials and contractors who have worked closely with the Pentagon specifically on AI.
A spokesman for Anthropic declined to comment when asked about the Pentagon's statement.
The details of the defence department stopping all use of Anthropic's tools have not been previously reported.
Anthropic 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.
But earlier this year, the Pentagon pressed it to remove its safety guardrails from Claude and give the military unfettered access to its AI tools.
When Anthropic refused, citing concerns of mass surveillance and autonomous weapons, Hegseth designated the company a "national security supply-chain risk" - a law typically reserved for companies based in countries that pose a threat to the US.
Anthropic said the move was "unprecedented and unlawful" and has sued the Trump administration in an effort to overturn the designation.
As this legal fight has played out in the courts, the Pentagon has 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 two people familiar with AI use in the defence department
The Pentagon's statement to the BBC on Monday came after multiple people said in interviews that Claude was still being widely used in recent weeks.
It was embedded in a larger data platform used by the Pentagon called Maven Smart System, multiple sources said.
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 said that because Claude was deeply integrated into Maven it would have been difficult to remove it quickly 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 said.
"Once they become integrated it can be painful to remove them," added Kahn, a former US defence official.
The BBC spoke with defence contractors who specialise in AI and have worked closely with the Pentagon, as well as former US defence officials and others familiar with the Pentagon's use of artificial intelligence. Several spoke on the condition of anonymity because they were not authorised to speak to the media.
When Hegseth initially ordered Anthropic to be phased out of government, "there was concern with moving beyond Anthropic to OpenAI" and other tools, a former senior US defence official said.
Hegseth set a six-month deadline to phase out Anthropic on 27 February, one day before the US and Israel attacked Iran.
Pentagon employees continued to use Anthropic's Claude tools, including its powerful Mythos models, "throughout the whole period of the controversy" with the company, one of the people said.
Claude was used in particular for intelligence gathering and analysis, including through Palantir's Maven system, several sources said. 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 used said analysts and other defence officials rely 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, according to a source who was given a demonstration of Maven by Palantir earlier this year.
Recent private demonstrations of Maven for government officials have also included Claude, another person familiar with Maven said.
The Maven system is deployed "across the entire department", Cameron Stanley, who leads AI work within the defence department, said during a demonstration of Maven earlier this year.
Maven is also used by some US allies to share intelligence with the Trump administration.
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 meeting.
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 said he still had concerns about risks posed by AI.
"How we address those risks is still under discussion," Amodei told reporters at the White House, as Trump stood by his side, but "if we work with the president, we can win safely".
Trump then proceeded to praise Amodei, saying he was "great" and "fantastic."
Sign up for our Tech Decoded newsletter to follow the world's top tech stories and trends. Outside the UK? Sign up here.