Pentagon under pressure to drop Anthropic’s Claude as military users push back

Despite a formal directive to ban Anthropic’s Claude from their systems, a coalition of Pentagon officials, staffers, and IT contractors are reportedly resisting.

As reported by Reuters, military users see Anthropic’s technology is significantly superior to available alternatives, sowing a division within the US military’s tech ranks.

Since February, Anthropic and the Pentagon have been locked in conflict over AI ethics and safety protocol while using the artificial intelligence tools for military purposes.

In early March, Defense Secretary Pete Hegseth designated the company a supply-chain risk, banning its use by the Pentagon and its contractors following a six-month phase-out.

The move came after Anthropic CEO Dario Amodei refused to give in to the Pentagon demands, forcing the tech company to remove guardrails for all lawful uses. However, Anthropic challenged the stance based on two reasons: mass surveillance of American people and managing lethal autonomous weapons.

“Career IT people at DoD hate this move because they have finally gotten operators comfortable using AI,” said one IT contractor.

The contractor said Anthropic’s Claude AI model “is the best, while xAI’s Grok often produced ​inconsistent answers to the same query.”

In July 2025, Anthropic became a part and parcel of the US military tech after securing a $200 million defense contract. Since then, AI tools have been used for various high-stakes tasks including targeting and operations, data analysis, and software development.

According to experts, moving away from Anthropic is not easy as it seems as it is mired with critical roadblocks.

What makes the transition a major undertaking is the integration of Claude-built multiple prompts and workflows into Palantir's Maven Smart ​Systems, a software platform that supplies militaries with ​intelligence analysis and weapons targeting, as confirmed by sources privy to the matter.

Joe Saunders, CEO of RunSafe Security, notes that replacing existing models requires a gruelling recertification process for classified use, which can take 12 to 18 months.

According to insiders, the tasks once handled by Claude are now completed manually by using Microsoft Excel, demonstrating the loss of efficiency associated with Anthropic Claude.

"What we are seeing play out here is the tension of adoption, both inside the Pentagon as well as the political level," said ​Roger Zakheim, director of the Ronald Reagan Presidential Foundation and Institute.