The Donald Trump administration ousted the artificial intelligence (AI) corporations Anthropic, putting AI data analytics corporations Palantir in a difficult position.
The Ministry of National Defense's AI-based military intelligence platform used in the U.S. strike on Iran, "Maven Smart System," was built by Palantir on Anthropic's "Claude Code," and because Anthropic was designated a supply chain risk corporations, Palantir has been forced to remove Claude from the platform.
Reuters reported this on the 4th (local time), citing multiple sources familiar with the matter. The "Maven Smart System," used to strike more than 1,000 targets in the first 24 hours of the U.S. strike on Iran, is a platform that analyzes vast military data, identifies targets, and supports rapid decision-making on the battlefield. According to the sources, the "Maven Smart System" uses Anthropic's developer AI coding tool "Claude Code" for its base prompts and workflows.
Earlier, the Ministry of National Defense and Anthropic clashed over how and to what extent AI should be used. The Ministry of National Defense demanded fully opening the scope of AI's military use, but Anthropic held to the position that its technology should not be used for large-scale surveillance or the development of fully autonomous lethal weapons. When Anthropic refused to accept the Ministry of National Defense's demands to the end, President Trump ordered all government agencies to stop using Anthropic's technology, and the Ministry of National Defense designated Anthropic a "supply chain risk" corporations. As a result, Anthropic has been barred from any commercial transaction with the tens of thousands of contractors that do business with the U.S. Ministry of National Defense.
Reuters said that, due to this measure, Palantir, which has contracts worth at least $1 billion (about 1.47 trillion won) with major government agencies including the U.S. Ministry of National Defense, has also been forced to replace the Claude embedded in the "Maven Smart System" with another AI model and to rebuild parts of the software. Because Anthropic's technology is deeply integrated across public and private systems, including the "Maven Smart System," a full replacement is expected to take considerable time and expense.
Anthropic has signaled legal action against the Ministry of National Defense's designation of it as a supply chain risk corporations, but it has not formally responded yet because the Ministry of National Defense has not begun procedures such as publishing the notice in the official gazette.
Meanwhile, the U.S. strike on Iran, Operation "Epic Fury," is drawing attention as the first war in which a large language model (LLM), which has advanced rapidly since the debut of ChatGPT in 2022, was officially deployed as the "brain" of the battlefield. AI was used in the Russia-Ukraine war, but this is the first time a general-purpose AI like Anthropic's Claude has been coupled to existing systems as a reasoning engine and employed in military operations.