U.S. President Donald Trump ordered federal government agencies to stop using Anthropic AI, but the U.S. military continued to use Anthropic's Claude AI in a large-scale air strike on Iran carried out immediately after the order, the Wall Street Journal reported on Feb. 28 local time.

The Wall Street Journal cited people familiar with the matter as saying major commands worldwide, including U.S. Central Command, which oversees the Middle East, are using Claude.

Military officials said Central Command has used Claude for intelligence analysis, target identification and combat scenario simulations.

The dispute between the Trump administration and Anthropic has continued for months. During contract talks, the Pentagon demanded it be allowed to use the tool in "all lawful scenarios," and Anthropic rejected the request.

The U.S. government also took issue with what it said were Anthropic's links to major Democratic Party donor groups and lobbying activities opposing the administration's AI policy. Trump ordered relevant agencies last Friday to stop cooperating with Anthropic, and the Defense Department designated Anthropic as a supply-chain security threat.

Claude was also used in a U.S. military operation that arrested Venezuelan President Nicolas Maduro.

The Defense Department signed contracts with OpenAI and xAI, led by Elon Musk, for use in classified environments to replace Claude. AI experts expect it will take months to fully replace Claude with other models, the Wall Street Journal reported. Claude is connected to systems at several partners, including data analytics company Palantir, making it difficult to replace quickly, the report said.