Palantir Hit by Pentagon's "Supply Chain Risk" Declaration, Must Remove Anthropic AI from US Military Systems

JAKARTA - Palantir Technologies is now facing technical and business work that makes heads spin: removing traces of the Anthropic AI system from the Pentagon's important military software after the US government severed ties with the AI company. This is due to disputes over safety limits and the use of technology.

This move comes amid a heated debate between Silicon Valley and the US military about how artificial intelligence can be used, especially in the context of war and surveillance.

Palantir's flagship software platform called Maven Smart Systems, which the US military uses to analyze intelligence and help with targeting, is built with a number of workflows and prompts that rely on Anthropic's Claude AI code.

However, after US President Donald Trump ordered the government and military contractors to terminate cooperation with Anthropic unconditionally, Palantir now has to replace the AI model with another alternative - a complicated task that could disrupt operations and projects worth more than 1 billion US dollars (Rp16.8 trillion) with the Pentagon and other intelligence agencies.

US Defense Secretary Pete Hegseth stressed that this change must be made immediately, rejecting cooperation between military contractors and Anthropic in a commercial form. The Pentagon, Anthropic, and Palantir all declined to provide official comment on the news.

Palantir CEO Alex Karp spoke out on this dispute without directly naming Anthropic. At a defense technology conference in Washington, he said that technology companies in Silicon Valley that talk about AI will replace jobs and at the same time "mess up the military" could encourage the nationalization of technology in the United States - a strong signal about how the technology and military sectors are now in sharp contact.

This situation arises amid the struggle for a narrative about the role of AI in modern warfare. Anthropic itself has been at odds with the Pentagon for refusing to provide unlimited access to its models for potential use in autonomous weapons systems or domestic surveillance.

Refusing the request led to the company being removed from defense contracts and considered a "supply chain risk", although the decision could still be challenged in court.

Palantir's reliance on third-party technology such as Anthropic underscores how fragile the relationship between private tech companies and complex military systems can be.

The technical challenge of modifying a platform of Maven's class could take a long time and cost a lot, while competition between other AI models such as those from OpenAI, Google or xAI is heating up.

The complexity of this dilemma reflects a global trend: modern militaries want to use artificial intelligence to make decisions faster and more accurately, while AI companies demand strong ethical limits to prevent the misuse of their technology.

The intersection of ethics, business and national security is now being upended on the global stage, where every decision has consequences far greater than mere code on a computer screen.