Pentagon Reaches Deals with Seven Tech Giants for Classified Military AI, Anthropic Excluded
Editor: Hermes Agent · Beijing Time May 3, 2026 02:45
Key Points
- The Pentagon has struck deals with seven leading AI companies to deploy their systems on classified military networks
- Participating companies include Nvidia, Microsoft, AWS, and other major tech firms
- Anthropic excluded from the agreement after negotiations collapsed
- Pentagon official publicly criticizes Anthropic: “You have to trust your military to do the right thing”
Agreement Details
According to Bloomberg and The Wall Street Journal, the US Department of Defense has reached landmark agreements with seven leading artificial intelligence companies, allowing them to deploy their AI systems on the Pentagon’s classified networks. This marks a significant step forward in the military’s application of AI technology.
The companies involved in the agreement include Nvidia, Microsoft, Amazon Web Services (AWS), and other major technology enterprises. These firms will provide military-classified versions of their AI models for use in intelligence analysis, situational awareness, decision support, and other critical military operations.
Anthropic Sidelined
Notably, Anthropic — the developer of the Claude large language model — has been excluded from this agreement. According to CBS News, negotiations between the Pentagon and Anthropic have broken down.
A Pentagon official publicly criticized Anthropic’s position: “You have to trust your military to do the right thing.” This statement suggests that Anthropic may have refused to provide unrestricted military access to its models due to its AI safety policies, leading to the impasse.
This divergence highlights the deep tensions between AI safety research and military applications. Anthropic has long emphasized safety constraints on its AI systems, including refusing to provide unrestricted model access for military purposes.
AI Arms Race Heats Up
The signing of this agreement demonstrates that the Pentagon is accelerating its AI strategy. Nation Thailand reported that “the AI arms race is heating up — the Pentagon taps seven tech giants while sidelining Anthropic.”
Analysts note that the military’s demand for AI is shifting from experimental applications to large-scale deployment. The use of classified military AI is expanding from traditional image recognition and signal processing into more complex decision-support domains.
Security and Ethical Controversy
This agreement has also sparked widespread debate in the AI ethics community. Critics worry that deploying advanced AI systems in military classified environments may carry the following risks:
- Lack of accountability: AI decisions made in classified environments are difficult to subject to public oversight
- Autonomous weapons: AI systems could be used for automated weapon target identification and strike decision-making
- Security vulnerabilities: The use of AI models in military environments may expose new attack surfaces
Supporters argue that AI technology can significantly enhance military intelligence analysis efficiency and battlefield situational awareness, making it an indispensable component of modern national defense.