Pentagon Embedded Anthropic AI in Iran Strike Operations Hours After Trump Administration Ban
Intelligence sources confirm that United States military commands deployed Anthropic's Claude AI model for operational support during a major air strike on Iran, occurring mere hours after President Donald Trump ordered federal agencies to halt use of the company's systems.
Military commands, including US Central Command (CENTCOM) operating in the Middle East, utilized Claude for intelligence analysis, potential target identification, and battlefield simulation capabilities, according to sources familiar with the matter cited by The Wall Street Journal. This deployment illustrates the deep integration of advanced AI systems into defense operations, persisting even as the administration moved to sever institutional ties with Anthropic.
On Friday, the Trump administration issued directives ordering federal agencies to cease collaboration with Anthropic and instructing the Defense Department to classify the company as a potential security risk. This administrative action followed the breakdown of contract negotiations, during which Anthropic refused to grant unrestricted military use of its AI systems for any lawful scenario requested by defense officials.
The contradiction between the ban and operational reality raises significant questions about the coherence of current AI governance frameworks. While political leadership moved to restrict Anthropic's access to government systems, operational military commanders continued leveraging the company's technology through existing channels.
Anthropic had previously secured a multiyear Pentagon contract valued at up to $200 million, part of a broader initiative involving multiple major AI laboratories. Through partnerships established with Palantir and Amazon Web Services, Claude became an approved system for classified intelligence and operational workflows. Prior to the recent Iran operation, Claude reportedly supported earlier missions including a January operation in Venezuela that resulted in the capture of President Nicolás Maduro.
Tensions between Anthropic and the Defense Department escalated after Defense Secretary Pete Hegseth demanded the company permit unrestricted military application of its models. Anthropic CEO Dario Amodei rejected these demands, characterizing certain proposed applications as ethical boundaries the company would not cross, even at the cost of significant government contracts.
In response, the Pentagon initiated contingency planning with alternative providers, ultimately reaching agreement with OpenAI to deploy its AI models on classified military networks.
During a Saturday interview, CEO Amodei articulated the company's position, stating that Anthropic opposes the use of its AI models for mass domestic surveillance and fully autonomous weapons systems. He emphasized that military decision-making should remain under human control rather than being delegated entirely to machines.
The incident highlights the operational complexities of rapid AI decoupling in defense contexts, where technical integration often precedes and outpaces policy decisions.