Palantir CEO Alex Karp stated that the company is still using Anthropic's Claude artificial intelligence models, despite the Pentagon's recent designation of Anthropic as a supply-chain risk and ongoing disputes between the startup and the Department of Defense (DOD) [1]. Karp explained that Palantir's products remain integrated with Anthropic, but suggested future integration with other large language models is likely [1]. The DOD officially labeled Anthropic a supply-chain risk last week, yet continues to use Claude models to support military operations in Iran [1].
Anthropic responded by suing the Trump administration on Monday, seeking to reverse the supply-chain risk designation and requesting a stay on the DOD action [1]. This marks Palantir's first public comment since the Pentagon's designation. Anthropic and Palantir previously partnered with Amazon Web Services to support the DOD in 2024 [1]. Other defense technology companies, including Lockheed Martin, have instructed employees to stop using Claude as the dispute unfolds [1].
Defense Department Chief Technology Officer Emil Michael told CNBC that transitioning away from Anthropic's models will take time, emphasizing that "you can't just rip out a system that's deeply embedded overnight" [1]. President Donald Trump, in a Truth Social post at the end of February, criticized Anthropic's staff and announced a six-month period for federal agencies to phase out the company's products [1]. However, an internal Pentagon memo from Chief Information Officer Kirsten Davies indicated that use of Anthropic's tools may continue beyond the six-month period if deemed critical to national security, with exemptions considered for "mission-critical activities" where "no viable alternative exists" [1]. Michael reiterated that exceptions would be made for sensitive operations to avoid risking current missions, but otherwise, the six-month phase-out remains the plan [1].
CONCLUSION
The Pentagon's designation of Anthropic as a supply-chain risk has triggered a high-profile dispute, with Palantir and other defense tech firms navigating the transition away from Anthropic's AI models. While a six-month phase-out is planned, critical operations may still rely on Claude, and legal action is underway. The situation is likely to have significant implications for defense technology providers and their partnerships.