Palantir Technologies confirmed it remains connected to Anthropic’s Claude artificial intelligence model while a dispute between Anthropic and the U.S. Department of Defense unfolds. The company’s chief executive, Alex Karp, made the comments during an onstage interview at Palantir’s AIPcon 9 in Maryland.
On Thursday, Karp told Seema Mody of CNBC that despite steps the Defense Department has discussed, Anthropic has not yet been removed from the department’s supply chain. "The Department of War is planning to phase out Anthropic; currently, it’s not phased out," he said. Karp added that Palantir’s products are currently integrated with Anthropic and indicated those integrations could broaden over time. "Our products are integrated with Anthropic, and in the future, it will probably be integrated with other large language models," he said.
The Pentagon formally designated Anthropic a supply-chain risk last week. At the same time, reporting has indicated the Defense Department continues to employ Claude models in support of operations related to the war in Iran.
The comments from Palantir’s CEO come at a public company conference focused on artificial intelligence and follow the DoD’s recent designation of Anthropic. Karp’s statements emphasize that, as of now, Anthropic remains part of Palantir’s technology stack and that Palantir expects to work with multiple large language models as its products evolve.
Key points
- Palantir continues to use Anthropic’s Claude model in its product integrations, according to CEO Alex Karp.
- The Department of Defense has designated Anthropic a supply-chain risk but reportedly still uses Claude to support the war in Iran.
- Palantir expects future integrations to include other large language models, indicating potential diversification of AI providers.
Sectors potentially affected - defense procurement, enterprise software and AI infrastructure.
Risks and uncertainties
- Timing and scope of any DoD phase-out of Anthropic remain uncertain - this affects defense contractors and suppliers reliant on Anthropic.
- Continuing operational use of Claude by the Defense Department, despite the supply-chain designation, creates ambiguity about procurement and compliance risk for defense and AI vendors.
- Palantir’s potential shift to include other large language models introduces uncertainty around integration work, compatibility and vendor concentration risk for its product ecosystem.
These developments highlight an ongoing intersection between defense procurement decisions and commercial AI deployments. Palantir’s public confirmation of current Anthropic integration and its openness to broader model support underscore continued reliance on third-party models while also signaling a pathway toward multi-vendor AI strategies.