WayToClawEarn
High impactThe Verge / CNBC

The Pentagon signs contracts with 7 AI companies: OpenAI, Google, Nvidia are included, Anthropic is excluded

The U.S. Department of Defense officially announced the signing of confidential-level AI use agreements with OpenAI, Google, Microsoft, Amazon, Nvidia, xAI, and Reflection, and also listed the previous partner Anthropic as a supply chain risk and excluded it. This is an important milestone in the militarization of AI.

WayToClawEarn EditorialPublished May 2, 2026Updated Aug 8, 2026

Editorial review of public sources · AI-assisted drafting. How we work · Original source

Core conclusion

On May 1, 2026, the U.S. Department of Defense (Pentagon) officially announced that it had signed agreements with a total of 7 AI companies including OpenAI, Google, Microsoft, Amazon, Nvidia, xAI (Elon Musk) and the startup Reflection, allowing it to use these companies' AI tools in a classified environment. Anthropic, meanwhile, was specifically excluded - the Department of Defense classified it as a "supply chain risk."

This move marks that the U.S. military has officially entered the "AI first" era, and it also means that the business model of AI companies will usher in a new dimension of large government-level customers.

Key Points

  • Event time: May 1, 2026
  • Signatories: OpenAI, Google, Microsoft, Amazon, Nvidia, xAI, Reflection
  • Excluded party: Anthropic (previously owned $2 million confidential contracts)
  • Core signal: "Military compliance" of AI companies will become a new competitive watershed

Background: From cooperation to division

Over the past two years, the relationship between AI companies and the U.S. military has experienced a roller coaster ride from honeymoon to breakup.

Anthropic has previously signed a $200 million contract with the Pentagon to handle classified material. However, Anthropic is adamant about relaxing its so-called "red lines" - refusing to make concessions for the Department of Defense on massive domestic surveillance and fully autonomous weapons systems. This led to a breakdown in the relationship, and the Department of Defense subsequently banned Anthropic's products from federal use.

Anthropic responded by taking the federal government to court and won a temporary injunction. The outcome of this lawsuit is still pending.

While Anthropic is in a stalemate with the Department of Defense, OpenAI and xAI have already taken the lead and reached agreements with the Pentagon on the "legal use" of their AI systems. Google has reached a similar agreement, The Information reported. Microsoft and Amazon were naturally selected because of their "deep foundation of cooperation with the Department of Defense."

Key Impact

DimensionsChangesImpactRecommended actions
AI industry landscape7 companies divide the defense market, Anthropic is marginalizedAI company valuations will diverge due to government contractsPay attention to API updates and compliance features of companies that sign the agreement
Military AI applicationsFrom experiments to "operational" useThe speed of AI implementation in military scenarios will be greatly increasedMilitary-grade AI security technology may spill over to the civilian market
Supply chain securityAnthropic is identified as a "supply chain risk"Compliance becomes a new competitive threshold for AI companiesAssess whether your own AI tool chain involves banned suppliers
Legal precedentMythos model hailed as "national security moment" by Defense Department CTOStrong AI security models can also be bargaining chipsAnthropic's Mythos model deserves attention

Adaptation suggestions

For AI content entrepreneurs

  • Watch for API updates from contracted companies: Being recognized by the military means these companies have the highest level of verification of security and compliance
  • Pay attention to xAI/Grok’s pricing strategy: Grok 4.3 is already one of the fastest and cheapest models on the market, and it may further expand its market share after receiving military endorsement
  • Anthropic’s Mythos Model: A cybersecurity model publicly praised by DoD CTOs that may represent the next wave of opportunities in AI security

For automation developers

  • The APIs of the 7 contracted companies will undergo the most stringent security review, which means that stability will be greatly improved
  • Military-grade AI security tools such as Mythos may be available to the civilian market within the next 6-12 months
  • Note compliance differences: some models are available in a military environment but may have different terms in a civilian environment

Tool entry

In this incident, AI tools from OpenAI, Google, Microsoft, Nvidia, xAI (Grok) and other companies have obtained the highest level of security certification. The Claude and Mythos models of Anthropic were excluded due to compliance disputes. API and model choices for these platforms will directly impact technology stack decisions for automated workflows.

Internal link guidance

Related

Related extended information

View source →

Disclaimer: this site shares educational insights only, for inspiration and reference. No outcome guarantee; external execution and decisions are your own responsibility.