The Pentagon has stopped using Anthropic's AI tools months after it labelled the company a 'supply chain risk', according to the BBC, which reports the firm's Claude model had recently been in wide use.
Key facts
- The Pentagon has stopped using Anthropic's AI tools, according to the BBC.
- Anthropic was labelled a 'supply chain risk' in February.
- The designation followed Anthropic's refusal to remove safety guardrails from its tools.
- Multiple sources told the BBC that Anthropic's Claude model was recently widely used.
The US Department of Defense has stopped using artificial intelligence tools made by Anthropic, months after it blacklisted the company, according to the BBC. The move follows the Pentagon’s decision earlier this year to label the firm a “supply chain risk”.
According to the BBC, that designation was made in February after Anthropic refused to remove safety guardrails from its tools. The company’s stance appears to have been a central factor in the Pentagon’s decision to flag it as a risk.
Despite the blacklisting, multiple sources told the BBC that Anthropic’s Claude model was recently widely used within the department. That reported continued use highlights a gap between the formal designation applied to the company and how its technology was actually being deployed.
The BBC reports that the Pentagon has now ceased using the Anthropic tools, bringing its practices in line with the earlier risk label. The development underscores the tensions that can arise between AI developers and large government customers over how such tools are built and safeguarded.
The sources cited by the BBC did not, in the information provided, detail the full scope of how Claude had been used or the reasons behind the decision to finally stop. The reporting centres on the timeline running from the February designation to the recent halt in use.
Why it matters
The episode illustrates the friction between AI companies that build in safety restrictions and powerful institutional buyers that may want those limits removed. It also shows how a formal risk designation does not always immediately translate into changes in how technology is used on the ground.
Frequently asked questions
Why did the Pentagon label Anthropic a supply chain risk?
According to the BBC, the Pentagon labelled Anthropic a 'supply chain risk' in February after the firm refused to remove safety guardrails from its tools.
Was Anthropic's AI still being used after the blacklisting?
Yes. Multiple sources told the BBC that Anthropic's Claude model was recently widely used, despite the earlier designation.
What has the Pentagon done now?
The BBC reports the Pentagon has stopped using Anthropic's AI tools, months after labelling the company a supply chain risk.