Key Points
- On March 3, Defense Secretary Pete Hegseth labeled Anthropic a “supply chain risk,” mandating a six-month removal period for Claude AI throughout Defense Department operations
- Defense technology personnel and contractors argue Claude outperforms competing systems like xAI’s Grok and oppose the transition
- Technical experts estimate the replacement process will require 12–18 months and impose significant costs in operational efficiency and system recertification
- Tech industry leaders including OpenAI, Amazon, Microsoft, and Google are actively lobbying Defense Department officials to overturn the classification
- Multiple government agencies are intentionally delaying implementation, anticipating a negotiated resolution before the enforcement deadline
Defense Secretary Pete Hegseth issued directives in March 2026 requiring the Pentagon to discontinue use of Anthropic’s Claude AI platform, triggering substantial resistance from military personnel and defense contractors.
Hegseth classified Anthropic as a “supply chain risk” on March 3, following disagreements regarding usage restrictions and safeguards for the company’s artificial intelligence capabilities in military applications. The directive prohibits both Pentagon operations and defense contractors from utilizing Anthropic’s products, establishing a six-month transition period for complete removal.
However, implementation is encountering significant obstacles. Defense Department information technology specialists, former Pentagon officials, and contracting firms express strong reluctance to discontinue Claude, citing its superior performance compared to existing replacement options.
“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 secured a $200 million defense contract in July 2025. Claude achieved the distinction of becoming the first artificial intelligence model authorized for deployment on classified military networks, experiencing widespread adoption throughout federal government operations.
Reuters previously documented that Pentagon operations utilized Claude in supporting United States military activities during confrontations with Iran. Intelligence sources indicate the technology remains operational despite the formal prohibition.
Replacement Challenges and Costs
Transitioning away from Claude involves complex technical challenges beyond simple system substitution. Joe Saunders, CEO of government defense contractor RunSafe Security, estimates that recertifying alternative systems for classified network deployment could require 12 to 18 months.
“It’s not just costly, it’s a loss of productivity,” Saunders said.
Operations previously automated through Claude, including large-scale data analysis and querying, are now being conducted manually using legacy tools like Microsoft Excel in certain instances. Claude Code, which gained extensive adoption within Pentagon software development operations, has created significant frustration among developers following its withdrawal.
Palantir’s Maven Smart Systems, a comprehensive platform deployed for military intelligence assessment and weapons targeting coordination, was constructed utilizing Anthropic’s Claude Code. Palantir maintains Maven-related contracts valued above $1 billion and faces substantial software reconstruction requirements using alternative AI platforms.
Several contractors are intentionally “slow-rolling” the transition, continuing to leverage Claude for workflow development ahead of the enforcement deadline.
Corporate Pressure Intensifies
Numerous technology corporations, including OpenAI, are discreetly pressuring Defense Department leadership to rescind the supply chain risk classification, according to reporting from the New York Times.
Principal Anthropic investors and strategic partners — including Amazon, Microsoft, and Google — maintain substantial financial interests in the company and are actively challenging the designation.
Industry executives express concerns that the Pentagon’s decision could establish a sweeping regulatory precedent influencing how government contractors engage with artificial intelligence providers.
One chief information officer at a federal agency confirmed plans to deliberately delay the phase-out timeline, anticipating successful negotiations between government officials and Anthropic before the six-month deadline arrives.
