Pentagon Halts Use of Anthropic AI After Months of National Security Dispute

The US Department of Defence has officially stopped using artificial intelligence tools developed by Anthropic, according to a statement given to the BBC, bringing an end to a contentious period in which the company's technology remained embedded in military systems despite being designated a national security risk earlier this year.

A department official confirmed that the Pentagon "has ceased the use of Anthropic products," marking the first public acknowledgment that all usage of the company's AI has been discontinued. The development follows months of tension between the defence establishment and one of the world's leading AI developers.

The dispute dates back to February, when Secretary of Defence Pete Hegseth classified Anthropic as a supply chain risk on national security grounds. At the time, Hegseth announced that the Pentagon would phase out the company's tools by late August, setting a six-month deadline that fell just one day before the US and Israel launched attacks on Iran. The designation is a legal mechanism typically reserved for companies headquartered in nations considered hostile to the United States.

Despite the phase-out order, multiple people familiar with the situation told the BBC that Anthropic's Claude model continued to be used by the department as recently as last week. The technology was reportedly employed for research, analysis, and intelligence gathering, as well as in military operations against Iran. Sources described Claude as a critical tool for processing and sorting through enormous quantities of information.

The people who spoke to the BBC included former US defence officials and contractors who have worked closely with the Pentagon on AI-related matters. They indicated that Claude was deeply embedded in the Maven Smart System, a data platform operated by Palantir that serves as the Pentagon's primary tool for organising intelligence and other information. Because of this integration, removing the technology quickly proved difficult.

The roots of the conflict lie in the Pentagon's request earlier this year that Anthropic remove safety guardrails from Claude and provide the military with unrestricted access to its AI capabilities. Anthropic declined, citing concerns about mass surveillance and the development of autonomous weapons systems. In response, Hegseth issued the supply chain risk designation, a move the company described as "unprecedented and unlawful."

Anthropic has since filed a lawsuit against the Trump administration in an effort to overturn the designation. The legal battle has unfolded alongside the Pentagon's efforts to diversify its AI partnerships, with contracts signed with Google, xAI, and OpenAI. According to two people familiar with AI use in the defence department, OpenAI's tools have been more widely adopted in recent months across some military departments.

Lauren Kahn, a senior research analyst at Georgetown's Center for Security and Emerging Technology, noted that Claude's deep integration into the Maven system made rapid removal impractical. "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. Kahn, a former US defence official, added that "once they become integrated it can be painful to remove them."

The BBC's sources described how analysts and defence officials relied on Maven to access and organise visual data, including satellite imagery and drone footage. This 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 received a demonstration of Maven by Palantir earlier this year.

A former senior US defence official said that when Hegseth initially ordered Anthropic to be phased out, "there was concern with moving beyond Anthropic to OpenAI" and other available tools. Pentagon employees continued to use Claude, including its powerful Mythos models, "throughout the whole period of the controversy," one person said.

Anthropic had been used by the US government and military since 2024, becoming the first advanced AI company to have its tools deployed in government agencies conducting classified work. The company's refusal to remove safety protocols has placed it at the centre of a broader debate about the role of artificial intelligence in military operations and the ethical boundaries of AI development.

The Pentagon's statement to the BBC on Monday came after multiple sources said in interviews that Claude was still being widely used in recent weeks. A spokesman for Anthropic declined to comment when asked about the Pentagon's statement. The details of the defence department's complete cessation of Anthropic's tools had not been previously reported.

The case has drawn attention to the growing reliance of military institutions on commercial AI systems and the complications that arise when those relationships sour. As the legal fight continues in the courts, the Pentagon's transition to alternative providers such as OpenAI, Google, and xAI signals a significant realignment in how the US military accesses and utilises artificial intelligence technology.