When the whole world turned its back on ChatGPT to sign a contract with the US military, it turns out Claude was already deeply embedded in the system. While Anthropic tried to distance itself from military use, its AI model continued to play a role where it mattered most. At the same time, OpenAI stepped in to fill the gap, taking over a space that had suddenly opened.
The situation highlights how quickly alliances shift in the AI world, and how difficult it is to separate advanced technology from real-world operations once it becomes part of critical systems.
Why Did The US Military Continue Using Claude Despite The Ban?
The Guardian reports that the US military used Anthropic’s AI model Claude during the joint US-Israel strikes on Iran, even though Donald Trump had ordered all federal agencies to stop using it just hours earlier.
The tool was reportedly used for intelligence gathering, selecting targets, and running battlefield simulations, making it a key part of the operation.
The continued use of Claude points to a larger issue. AI systems like this are not easily removed once they are integrated into military workflows. Defence officials themselves admitted that the technology is deeply embedded, and removing it quickly is not practical.
Trump strongly criticised Anthropic, calling it a “Radical Left AI company run by people who have no idea what the real World is all about”.
The dispute had already been building after Claude was used in a January mission to capture Venezuela’s president, Nicolás Maduro, which violated the company’s usage policies.
What Led To The Fallout Between Anthropic And The Pentagon?
The relationship between Anthropic and the US government has been tense for months. The company objected to the use of its AI in military actions, as its terms clearly prohibit use for violence, weapons development, or surveillance.
Defence Secretary Pete Hegseth accused Anthropic of “arrogance and betrayal” and pushed for full access to its AI systems. At the same time, he acknowledged that cutting ties immediately would be difficult, allowing a transition period of up to six months.
As the gap widened, OpenAI moved in. CEO Sam Altman confirmed that the company had reached an agreement with the Pentagon to provide its tools, including ChatGPT, for classified use.
