Asger Risom
The US Department of Defense is reportedly considering cutting ties with Anthropic and labeling it a “supply chain risk” after clashes over how its AI tools may be used by the military, escalating tensions between Silicon Valley and the Pentagon.
One of the most powerful relationships in artificial intelligence may be unraveling.
The US Department of Defense is reportedly weighing a dramatic step that could reshape Anthropic’s future.
According to Axios, the Pentagon is considering severing business ties with the AI firm and formally designating it a “supply chain risk.” Such a label would not only end direct cooperation but could block other defense contractors from using Anthropic’s technology.
The move follows months of negotiations over how the company’s Claude AI models can be deployed in military settings.
Clash over control
Anthropic CEO Dario Amodei has drawn firm boundaries around the use of the company’s systems, arguing they should not power weapons that operate without human oversight or be used to spy on Americans.
Axios reported that the Defense Department wants AI providers to allow their technology to be used for “all lawful purposes,” a position that appears to conflict with Anthropic’s restrictions.
Claude is already embedded in certain military systems and is currently the only AI model available in some classified environments, according to the report. The Pentagon is also said to be negotiating terms with Google, Meta and xAI.
High-stakes leverage
While losing a Defense Department contract alone may not significantly damage Anthropic’s finances, a supply chain risk designation could carry broader consequences.
Companies working with the Pentagon could be forced to drop Anthropic’s tools, potentially affecting commercial partnerships beyond defense.
The San Francisco-based startup recently disclosed a $14 billion annual revenue run rate and has raised $30 billion in funding. With eight of the Fortune 10 listed among its customers, it is widely viewed as a leading IPO candidate.
A defining test
The standoff underscores a deeper tension between private AI developers and national security institutions.
For the Pentagon, access to advanced AI systems is increasingly strategic. For companies like Anthropic, public commitments to ethical limits are central to their brand and governance model.
If the Defense Department follows through, the outcome may signal how much control AI companies can realistically retain once their systems become critical to state power.
Source: Axios
