The U.S. Department of Defense is advancing a sweeping push to embed artificial intelligence across military operations, announcing eight new expanded partnerships with leading American technology companies that aims to reposition the U.S. military as an “AI-first” fighting force. The agreements, finalized Friday, bring Google, OpenAI, Amazon Web Services, Microsoft, SpaceX, Oracle, Nvidia, and emerging startup Reflection into the Pentagon’s growing AI ecosystem, clearing the way for these firms’ AI tools to be used for any lawful military and operational purpose.
In a public statement following the announcement, Pentagon officials emphasized that the multi-vendor strategy is designed to avoid overreliance on a single technology provider, a vulnerability commonly referred to as “vendor lock-in.” By tapping into a diverse range of AI capabilities built across the robust U.S. technology sector, defense leaders say warfighters will gain access to cutting-edge tools to respond faster to threats and protect national security. The department also highlighted early successes from its existing military AI platform, launched last year: more than one million defense personnel across the department have already adopted the platform, cutting processing time for many critical tasks from months to just days.
The announcement comes amid a high-profile public and legal split with leading AI developer Anthropic, which is notably absent from the new set of contracts. The San Francisco-based firm, which was the first AI company to deploy its models for U.S. classified government work, still has its Claude chatbot tools in use across multiple defense and civilian agencies. However, the relationship collapsed earlier this year after Anthropic CEO Dario Amodei publicly raised ethical alarms over the potential misuse of powerful AI, warning that defense agencies could use the technology to carry out mass domestic surveillance and deploy fully autonomous lethal weapons. The company refused to agree to contract language allowing “any lawful use” of its AI tools for military purposes.
In response, Defense Secretary Pete Hegseth labeled Anthropic a “supply chain risk,” barring the firm from new government work. Anthropic has filed a lawsuit against the federal government alleging unlawful retaliation for its ethical stance, with the case scheduled to go to court in September.
The rift between Anthropic and the Pentagon has opened new opportunities for competing AI firms to deepen their ties to the U.S. military. OpenAI, maker of the ChatGPT large language model, was the first to capitalize on the shift, finalizing its own contract with the Pentagon in late February. A company spokesperson framed the deal as a commitment to equipping U.S. defense personnel with the world’s most advanced tools, noting Friday’s announcement was simply a formalization of the existing agreement.
For Google, the partnership marks a milestone: while the company’s Gemini chatbot is already used across some civilian government agencies, this contract will clear Gemini to handle classified defense work for the first time. The expansion has already sparked internal pushback: earlier this week, hundreds of Google employees, including dozens of researchers from the company’s leading AI research arm DeepMind, sent an open letter to CEO Sundar Pichai urging the company to abandon the deeper military partnership. Google has not yet issued a public response to the request or the contract announcement.
Other partners bring unique AI capabilities to the new framework. SpaceX, led by Elon Musk, is the parent company of xAI, the startup behind the controversial Grok AI chatbot. While xAI’s model is widely seen as less technically advanced than offerings from Anthropic, OpenAI, and Google, the addition of SpaceX extends the Pentagon’s access to Musk’s sprawling aerospace and technology ecosystem. Nvidia, a leading producer of AI computing hardware, will contribute its open-source Nemotron large language model, while startup Reflection will offer its open-source Reflection 70B model; neither firm will provide hardware as part of the current agreement. Longtime government cloud providers Amazon Web Services, Microsoft, and Oracle will continue to host the defense department’s online AI infrastructure, expanding their existing services to accommodate the growing volume of AI models and tools. None of the firms—SpaceX, Nvidia, Reflection, Amazon, Microsoft, or Oracle—have responded to requests for comment on the new contracts.
Defense Secretary Hegseth has made accelerating AI adoption across the U.S. military a top priority, arguing that access to advanced AI has become a core determinant of military success in modern conflict. For years, the Pentagon has worked to build out its AI capabilities, and Friday’s announcement represents the most significant expansion of those efforts to date.