Palantir Technologies Inc. is under pressure to remove artificial intelligence tools developed by Anthropic from a crucial US military software platform, following a US government directive ordering contractors to cease cooperation with the AI startup.

According to informed sources, Palantir's Maven Smart System platform, which assists military agencies in analyzing intelligence data and supporting weapons targeting, reportedly relies on prompts and workflows built using Anthropic's Claude AI model.

This challenge stems from an order by US President Trump instructing government contractors to halt collaboration with Anthropic. The company has been in conflict with the Pentagon over safety protocols related to autonomous weapons and surveillance technologies.

Consequently, Palantir may need to replace the Claude model with an alternative AI and reconstruct parts of its Maven platform. One source indicated that while US Secretary of Defense Pete Hegesses has demanded immediate compliance, the transition process could take several months.

Palantir holds contracts related to the Maven platform with the US Department of Defense and other national security agencies, with a potential total value exceeding $1 billion.

Other defense contractors, such as Lockheed Martin, may also be required to remove Anthropic's AI tools from their systems. This could lead to broader disruptions as government entities sever ties with the AI company.