The Department of Defense is currently testing AI solutions from OpenAI, Google, and xAI as potential replacements for Anthropic's Claude, which has been effectively excluded from military applications due to its refusal to lift certain safety restrictions. This competitive evaluation began on March 1, 2026, shortly after Defense Secretary Pete Hegseth identified Anthropic’s AI as a supply-chain risk.
Concerns regarding Anthropic’s technology arose when the company resisted demands for unrestricted use of Claude in sensitive areas, such as mass surveillance and lethal autonomous weaponry. This standoff intensified after reports in January 2026 revealed that Claude had been used in intelligence analysis and operational strategies related to Iran. The friction escalated, leading to Hegseth’s designation on February 27, which effectively initiated the search for a new AI solution.
The $200 Million Contract Unraveled
In July 2025, Anthropic secured a contract worth up to $200 million to integrate Claude into classified military networks. However, tensions emerged when Pentagon officials pushed for broader use of the AI, particularly in national security contexts. Anthropic’s leadership firmly opposed these requests, creating a significant divide that ultimately led to the contract's collapse.
With tensions escalating, the Pentagon's decision to seek alternatives became essential. The evaluation process is now being conducted through GenAI.mil, a newly established platform that operates separately from the Pentagon's existing Maven Smart System. This initiative involves 25 designated military personnel who are rigorously testing the competing AI models to assess their capabilities in workflows that previously relied on Claude.
A New Era of Military AI
The models currently under scrutiny include those developed by OpenAI and Google, along with Grok from Elon Musk's xAI. The Pentagon’s exploration of these alternatives reflects a significant shift in its approach to AI technologies, prioritizing both operational effectiveness and adherence to ethical standards.
As military workflows increasingly depend on advanced AI, the implications of this transition are substantial. The outcomes of these evaluations will not only shape the future of military AI applications but may also influence broader discussions about the ethical use of AI in sensitive areas. The Department of Defense recognizes the importance of balancing innovation with caution, particularly in an era defined by rapid technological advancements. The replacement of Claude may mark a new chapter in how the military engages with AI, shaping future contracts and partnerships.
Quick answers
Why was Anthropic’s Claude blacklisted by the Pentagon?
Claude was blacklisted due to Anthropic's refusal to remove safety restrictions that limited its use for mass surveillance and lethal autonomous weaponry.
What is the value of the contract Anthropic lost?
Anthropic lost a contract valued at up to $200 million that was intended for integrating Claude into military networks.
Which companies are competing to replace Claude?
The companies competing to replace Claude include OpenAI, Google, and xAI's Grok.
How many personnel are involved in testing the new AI models?
Twenty-five military personnel are involved in testing the competing AI models.
The stories that move AI & crypto markets — before the market reacts.
Free. 7am ET. Five stories. 62,400 readers.
