According to sources familiar with the matter, U.S. AI developer Anthropic is currently in discussions with the Pentagon over a critical contract. The two parties are at an impasse due to concerns over safeguards for the use of the technology, with the main point of contention being whether Anthropic’s technology can be used for autonomous weapon deployment and surveillance.
Is the $200 Million Contract in Jeopardy?
In July 2025, the Pentagon made a high-profile announcement that it had signed cooperation agreements with four top AI companies—Anthropic, Google (NASDAQ:GOOGL), OpenAI, and xAI—each valued at $200 million. These contracts aim to leverage cutting-edge AI technology to address national security challenges and accelerate the deployment of AI in the military sector.
Recent developments reveal that Anthropic is seeking to add restrictions to the contract, explicitly prohibiting the U.S. government from using its AI technology, Claude, for autonomous weapon systems and domestic surveillance activities.
Clearly, Anthropic’s stance on the use of its AI technology conflicts with that of the Pentagon, further deepening the divide between the company and the Trump administration.
In response, Anthropic stated, “We are having in-depth discussions with the Department of Defense about how to proceed with this work.” However, sources familiar with the negotiations revealed that lengthy talks between the Pentagon and Anthropic have been ongoing, and they are still at a standstill.
This contract dispute could serve as a bellwether for the industry, as more AI companies will likely set clear boundaries for the application of their technology when collaborating with the military. Privacy protection is expected to become a core prerequisite for AI’s militarization.
Notably, this dispute comes at a sensitive time for Anthropic. The startup is preparing for its eventual public listing. Additionally, the company has invested significant resources in securing business within the U.S. national security sector and is attempting to play an active role in shaping government AI policies.