Palantir Technologies continues to operate its software with Anthropic's Claude AI models, even as the Pentagon reviews the startup's status as a potential supply-chain risk. Chief Executive Alex Karp clarified the situation during an interview at the company's AIPcon event in Maryland.

"The Department of Defense has announced a plan to phase out Anthropic, but that process is not complete," Karp told CNBC. "Our products currently work with Anthropic, and they will likely connect with other large language models in the future."

The Defense Department formally categorized Anthropic as a risk last week, citing concerns over its technology sourcing. Despite this designation, reports indicate Claude models remain in active use supporting military operations related to Iran.

Karp's comments signal that Palantir, a major defense contractor, is maintaining its existing technical integrations while preparing for a possible shift. The company's platforms are designed to work with various AI systems, suggesting a strategic flexibility as government procurement policies evolve. The situation remains fluid as the Pentagon's planned phase-out proceeds.

Source: CNBC