Pentagon's Ultimatum to Anthropic
The U.S. Department of Defense has issued a formal ultimatum to the artificial intelligence company Anthropic, demanding it revoke its internal prohibition on the use of its technology for autonomous weaponry. U.S. Secretary of Defense Pete Hagset has given Anthropic's CEO, Dario Amodei, a deadline of February 27, 2025, to comply. This demand carries significant weight because Anthropic is the sole provider of AI models operating within the Pentagon's classified networks, including SIPR and JWICS. The company holds a Department of Defense contract valued at approximately $200 million, underscoring its critical role in U.S. national security infrastructure.
Anthropic's Technology and Its Challenges
While Anthropic permits the use of its technologies for offensive cyber operations, it maintains strict bans on mass surveillance of U.S. citizens and the combat application of autonomous weapons. These two prohibitions have become central points of contention in the company's negotiations with the Pentagon. The establishment of the Chief Digital and Artificial Intelligence Office (CDAO) in March 2023 highlights the growing strategic focus on AI within the U.S. defense establishment.
Amidst this pressure, Anthropic is also contending with large-scale industrial espionage campaigns. The company has identified Chinese AI labs—DeepSeek, Moonshot, and MiniMax—as being behind efforts to illicitly extract capabilities from its Claude model to enhance their own systems. According to Anthropic, these labs generated over 16 million queries to Claude using roughly 24,000 fraudulent accounts. Senior officials have suggested that a new DeepSeek model was likely trained on Nvidia Blackwell chips, whose export to China is banned, and which may be housed in a DeepSeek data center in Inner Mongolia.
Anthropic's contract and technological position place it in a uniquely influential market role. As expert Greg Allen noted, the company's technologies are already integrated into combat missions and utilized for critical operations. However, a potential contract termination by the Pentagon could have severe financial repercussions for Anthropic, emphasizing the high stakes of maintaining its partnership with the U.S. military. This standoff reflects a broader, global debate on the ethical governance of AI in warfare, where the U.S. seeks to maintain a technological edge while navigating complex moral and legal frameworks.