TLDR
- Defense Secretary Pete Hegseth designated Anthropic a “supply chain risk” on March 3, ordering a six-month phase-out of Claude across the Pentagon
- Military IT staff and contractors say Claude is superior to alternatives like xAI’s Grok and are resisting the switch
- Replacing Claude could take 12–18 months and cost heavily in productivity and recertification
- Companies including OpenAI, Amazon, Microsoft, and Google are lobbying the DOD to reverse the designation
- Some federal agencies are deliberately slow-rolling the phase-out, betting a deal will be reached before the deadline
Defense Secretary Pete Hegseth ordered the Pentagon to stop using Anthropic’s Claude AI in March 2026, but military staff and contractors are pushing back hard against the decision.
Hegseth wants Pentagon to dump Anthropic's Claude, but military users say it's not so easy
More Here → https://t.co/UWVr1KNIDQ pic.twitter.com/FSl6Mq5avZ
— PiQ (@PiQSuite) March 19, 2026
Hegseth designated Anthropic a “supply chain risk” on March 3, after a dispute over guardrails on how the military could use the company’s AI tools. The order bars the Pentagon and its contractors from using Anthropic products, with a six-month window to phase them out.
But the phase-out is not going smoothly. Pentagon IT staff, former officials, and contractors say they are reluctant to abandon Claude, which they view as better than available alternatives.
“Career IT people at DoD hate this move because they had finally gotten operators comfortable using AI,” said one IT contractor. “They think it’s stupid.”
The same contractor said Claude “is the best,” while xAI’s Grok often gave inconsistent answers to the same question.
Anthropic signed a $200 million defense contract in July 2025. Claude became the first AI model approved to run on classified military networks, and adoption was strong across the federal government.
Reuters previously reported that the Pentagon used Claude to support U.S. military operations during the conflict with Iran. Sources say the technology is still in use despite the blacklisting.
The Cost of Switching
Replacing Claude is not a simple swap. Joe Saunders, CEO of government contractor RunSafe Security, said recertifying replacement systems for use on classified networks could take 12 to 18 months.
“It’s not just costly, it’s a loss of productivity,” Saunders said.
Tasks once handled by Claude, like querying large datasets, are now being done manually using tools like Microsoft Excel in some cases. Claude Code, used widely inside the Pentagon for writing software, has left developers frustrated after its removal.
Palantir’s Maven Smart Systems, a platform used for military intelligence analysis and weapons targeting, was built using Anthropic’s Claude Code. Palantir holds Maven-related contracts worth more than $1 billion and will have to rebuild parts of its software using a different AI model.
Some contractors are “slow-rolling” the replacement, using Claude to build out workflows before the deadline hits.
Industry Pushback Grows
Several tech companies, including OpenAI, are quietly urging the DOD to walk back the supply chain risk label, according to the New York Times.
Major Anthropic investors and partners — including Amazon, Microsoft, and Google — have financial ties to the company and are pushing back on the designation.
Executives across the industry fear the Pentagon’s move could set a broad precedent affecting how government contractors interact with AI companies.
One chief information officer at a federal agency said it plans to slow-roll the phase-out, betting the government and Anthropic will reach a deal before the six-month deadline expires.





