Key Takeaways
- Seven technology firms received Pentagon AI authorization: Nvidia, Microsoft, Amazon Web Services, SpaceX, OpenAI, Google, and Reflection AI
- Anthropic remains barred following its designation as a supply-chain security concern
- Defense Department personnel express frustration over losing access to Anthropic’s capabilities, which many consider best-in-class
- Contracts enable AI deployment on the military’s most restricted networks, Impact Levels 6 and 7
- More than 1.3 million Defense Department users have engaged with the GenAI.mil system within its first five months
The Defense Department revealed Friday that it has executed agreements with seven major technology providers to integrate cutting-edge artificial intelligence capabilities into its highest-security military networks. The partner companies include Nvidia, Microsoft, Amazon Web Services, SpaceX, OpenAI, Google, and Reflection AI.
These partnerships authorize the firms to deliver services within the Pentagon’s Impact Level 6 and 7 environments—the highest classification tiers maintained by U.S. defense operations.
Friday’s announcement also represented the Pentagon’s first public acknowledgment of its collaboration with Google, which news outlets had previously disclosed earlier in the week.
Amazon Web Services completed its contractual arrangement late Thursday evening, according to two Pentagon sources with knowledge of the negotiations.
Notably absent from the roster is Anthropic. The company was designated a supply-chain security risk by the Pentagon earlier this year, resulting in a prohibition on its technologies across the Department of Defense and affiliated contractors. Personnel have received directives to phase out Anthropic systems over a six-month timeline.
Despite the official restriction, Defense Department employees, former officials, and information technology contractors told Reuters they’re hesitant to abandon Anthropic’s platforms. Multiple sources characterize the company’s offerings as technically superior to competing options.
The Rationale Behind Anthropic’s Exclusion
In a Friday interview with CNBC, Pentagon Chief Technology Officer Emil Michael reaffirmed that Anthropic continues to pose supply-chain vulnerabilities. Michael also discussed Anthropic’s Mythos AI model, which has attracted scrutiny for its sophisticated cybersecurity features, describing it as representing “a separate national security moment.”
Mythos has been distributed to multiple organizations and companies to bolster defenses against digital threats. Whether the Pentagon has obtained access to the preview release remains unconfirmed.
President Donald Trump indicated last week that Anthropic was making progress in gaining favor with his administration. The statement suggested the possibility that the company’s Pentagon exclusion could eventually be reconsidered.
AI Integration Across Military Operations
The Pentagon’s GenAI.mil initiative has attracted more than 1.3 million Defense Department users during its initial five-month operational period.
Military personnel leverage AI tools across mission planning, supply chain management, target identification, and additional functions to enhance efficiency in complex operations.
The Pentagon indicated that broadening its roster of AI vendors will mitigate the risk of “vendor lock”—terminology referencing its previous heavy dependency on Anthropic’s platforms.
The Department of Defense’s official statement characterized the agreements as advancing “the transformation toward establishing the United States military as an AI-first fighting force.”
The multi-vendor strategy emerges as the Pentagon pursues diversification of its artificial intelligence supply chain and seeks to minimize reliance on individual providers.
The latest milestone arrived with Amazon Web Services completing its contract late Thursday, mere hours ahead of the Pentagon’s formal Friday disclosure.





