With Anthropic on its way out of the Pentagon, two of its biggest rivals are moving fast to fill the gap. Elon Musk's xAI and OpenAI both signed agreements last week to work on classified government systems, positioning themselves as potential replacements for Claude in the U.S. military's AI infrastructure, as reported by "Hvylya", citing The Washington Post.
The opening appeared after President Trump banned all government agencies from using Anthropic's tools following a dispute over the company's restrictions on military applications. The ban gave the Pentagon six months to transition away from Claude - the AI tool embedded in the Maven Smart System that has become central to military operations, including the ongoing campaign in Iran.
The stakes are enormous. Maven, built by Palantir, processes classified data from satellites, surveillance, and intelligence sources to provide real-time targeting and analysis. Over 20,000 military personnel across the armed forces use the system. Replacing Claude - the AI engine at its core - without disrupting ongoing operations will be a significant technical challenge.
Anthropic was the first major AI company to work with classified data, gaining a substantial head start in defense AI. CEO Dario Amodei said Claude was "extensively deployed" across the Defense Department and other security agencies, used for intelligence analysis and operational planning. That deep integration is precisely what makes the transition so complex.
Ben Van Roo, CEO of defense software startup Legion Intelligence, said that in his experience integrating generative AI into military systems, "the baseline use case is chat and advanced search functions - essentially summarizing information." Whether xAI or OpenAI can match Claude's performance in more advanced applications - like the real-time targeting used in Iran - remains an open question.
Also read: "Rules Are for Suckers": Nobel Laureate Compares Trump's Iran Strike to Mugabe's Rigged Lottery