Pentagon Anthropic AI Tools: US Defense Department Ends Use After Blacklisting
The US Department of Defense has stopped using Anthropic’s artificial intelligence tools, months after the Pentagon designated the company a national security supply-chain risk.
A Defense Department official confirmed to the BBC that the Pentagon had “ceased the use of Anthropic products.”
The decision marks the latest development in a major dispute between the US military and Anthropic, the company behind the Claude artificial intelligence system.
Pentagon Ends Use of Anthropic AI
The Pentagon’s decision became public on Monday, October 5, after the BBC asked the department about its continued use of Anthropic technology.
Defense Secretary Pete Hegseth had ordered the department to phase out Anthropic products earlier this year.
However, people familiar with the Pentagon’s AI operations told the BBC that Claude was still being used as recently as last week.
The sources included former US defense officials and contractors who have worked directly with the department’s artificial intelligence systems.
Claude Was Used for Military Work
Anthropic’s Claude models had become an important tool for some Pentagon activities.
People familiar with the technology said Claude was used for research, intelligence gathering and analysis. It was also used during US military operations involving Iran.
The AI system was integrated into the Pentagon’s Maven Smart System, a major platform used to organise and analyse intelligence.
Maven is operated by Palantir and can process large amounts of information, including satellite images and drone footage.
The information could then be examined with artificial intelligence to help officials identify patterns and prepare intelligence reports.
Why Did the Pentagon Blacklist Anthropic?
The dispute began after the Pentagon demanded broader access to Anthropic’s AI models.
The Defense Department wanted the company to remove restrictions that limited how its technology could be used by the military.
Anthropic refused to remove its safeguards. The company said it had concerns about uses such as mass surveillance and fully autonomous weapons.
Hegseth subsequently designated Anthropic a national security supply-chain risk in February.
He also announced that the Pentagon would stop using the company’s products by the end of August.
Anthropic called the decision unprecedented and unlawful and later sued the Trump administration.
Why Did It Take Months to Stop Using the Tools?
Although the Pentagon announced plans to remove Anthropic’s technology months ago, Claude continued to be used.
One reason was the deep integration of the AI system into existing military platforms.
Lauren Kahn, a senior research analyst at Georgetown’s Center for Security and Emerging Technology, told the BBC that removing Claude from systems such as Maven was not a simple process.
The technology had become part of wider workflows used by defense personnel. Replacing it required changes to systems and operations.
The continued use also reflected the importance of Claude to some Pentagon users.
Pentagon Turns to Other AI Companies
During the dispute with Anthropic, the Pentagon expanded relationships with other major artificial intelligence companies.
The Defense Department has signed agreements involving companies including OpenAI, Google and xAI.
People familiar with the department’s technology use told the BBC that OpenAI tools have become more widely used in some military departments in recent months.
The shift means the Pentagon now has to manage its AI needs while reducing its dependence on Anthropic’s technology.
Anthropic and the US Government
Anthropic’s dispute with the Pentagon comes despite the company’s continued engagement with the US government.
Anthropic’s chief executive, Dario Amodei, recently met President Donald Trump at the White House.
Amodei has continued to raise concerns about the risks associated with increasingly powerful AI systems.
At the same time, Anthropic has been seeking closer relationships with government agencies and preparing for major growth in the commercial AI market.
The company has argued that its safeguards are designed to prevent dangerous uses of its technology.
What Happens Next?
The Pentagon’s decision to end the use of Anthropic AI tools closes one stage of the dispute, but the broader disagreement is not over.
Anthropic is challenging the government’s supply-chain risk designation in court. The outcome could affect how the US government classifies and works with advanced AI companies.
The Pentagon must also ensure that replacing Claude does not disrupt military intelligence and analysis systems that previously relied on the technology.
For now, the Defense Department says it has ended its use of Anthropic products.
The move highlights the growing importance of artificial intelligence in national security and the difficult questions surrounding how powerful AI systems should be used by the military.
About the Author
