Pentagon Deployed Anthropic's AI in Iranian Attack Just Hours Following Trump's Federal Ban: Report
According to reports, the Pentagon utilized Claude AI from Anthropic for targeting operations and intelligence assessment during an Iranian military strike, occurring mere hours following Trump's directive banning federal use of the company's technology.
According to reports, American defense forces deployed Anthropic's technology during a significant aerial assault against Iran, just hours following President Donald Trump's directive ordering federal entities to cease utilizing the firm's platforms.
Various military command centers, particularly US Central Command (CENTCOM) operating in Middle Eastern territories, deployed Anthropic's Claude artificial intelligence platform for operational assistance, based on accounts from individuals with knowledge of the situation referenced by The Wall Street Journal. The technology has purportedly provided support in analyzing intelligence data, pinpointing prospective targets and executing simulations of combat scenarios.
The occurrence demonstrates the extent to which sophisticated artificial intelligence platforms have integrated themselves within defense-related operations. Despite the administration's efforts to disconnect relationships with the firm, Claude continued functioning as part of military operational procedures.
During Friday, the Trump administration issued instructions to government entities directing them to cease collaboration with the organization and commanded the Defense Department to classify it as a possible security threat. The directive emerged following the collapse of contractual negotiations, during which Anthropic declined to provide unrestricted military access to its artificial intelligence for any legitimate scenario demanded by defense authorities.
Anthropic's Claude AI used for classified operations
Anthropic had earlier obtained a multi-year Defense Department agreement valued at as much as $200 million in conjunction with multiple prominent AI laboratories. Via collaborative arrangements involving Palantir and Amazon Web Services, Claude received authorization for sensitive intelligence operations and classified operational processes. The platform was purportedly engaged in previous operations as well, including a mission conducted in January within Venezuela that led to the apprehension of President Nicolás Maduro.
Friction escalated following Defense Secretary Pete Hegseth's insistence that the organization authorize unrestricted military deployment of its platforms. Anthropic CEO Dario Amodei turned down the demand, characterizing specific applications as moral limits the organization refused to breach, regardless of potentially forfeiting government contracts.
Following this development, the Pentagon initiated efforts to secure alternative service providers, finalizing an arrangement with OpenAI to implement its artificial intelligence platforms across classified military infrastructure.
Anthropic CEO pushes back on Pentagon ban
In a Saturday interview, Anthropic CEO Dario Amodei stated the organization stands against utilizing its artificial intelligence platforms for widespread domestic monitoring and completely autonomous weaponry, addressing a US government order that designated the company a defense "supply chain risk" and prohibited contractors from deploying its offerings.
He contended that specific use cases violate core principles, underscoring that military judgments must stay under human oversight instead of being transferred completely to automated systems.