Pentagon Signs AI Contracts with Eight Tech Firms to Build AI-First Military

The Pentagon announced eight new contracts with leading technology companies to accelerate the integration of artificial intelligence into U.S. military operations, stating the agreements will transform the military into an AI-first fighting force. These deals involve Google, OpenAI, Amazon, Microsoft, SpaceX, Oracle, Nvidia, and the startup Reflection, allowing their AI tools for any lawful operational use.

The Pentagon emphasized that partnering with multiple firms prevents vendor lock-in and provides diverse capabilities to warfighters.

More than a million defense department personnel have used the military's AI platform since its launch last year, reducing task times from months to days. For instance, the Army XVIII Airborne Corps shortened operations order creation from nine to 12 months to about six weeks using generative AI.

The agreements build on existing relationships, such as OpenAI's deal signed in February and Google's Gemini now handling classified government work for the first time. SpaceX, through its xAI subsidiary, will provide tools like the Grok chatbot, while Nvidia and Reflection contribute open-source models Nemotron and Reflection 70B.

Microsoft, Amazon Web Services, and Oracle will continue providing cloud services tailored for government and military deployment of AI models.

Anthropic was notably absent from the contracts after refusing terms allowing any lawful use, citing concerns over potential applications in warfare and domestic surveillance. The company is suing the government, alleging retaliation when the Pentagon labeled it a supply chain risk in March, a designation that barred its tools from certain settings.

Anthropic's CEO Dario Amodei publicly expressed fears about AI enabling mass surveillance and autonomous weapons. Despite the fallout, Anthropic's Claude model remains in use in some government agencies, and talks continue with the White House about a possible new deal.

leadership amid global competition, particularly with China. A recent White House memo highlighted foreign attempts, mainly from China, to steal U.S. AI systems, prompting efforts to share information and develop defenses. Defense officials affirmed that AI will not make lethal decisions, with humans retaining control, in line with legal requirements.

The strategy includes accelerating intelligence analysis and AI-aided simulations to enhance decision-making and domain awareness. Earlier this year, hundreds of Google employees urged their company not to deepen government ties, in a letter to CEO Sundar Pichai.

The Pentagon's push includes tools like Project Maven, which uses AI to synthesize intelligence data for targeting, as demonstrated in operations involving real-time satellite and surveillance inputs.

The U.S. is intensifying AI adoption to counter advancements by adversaries, such as China's autonomous drones displayed in a recent military parade. Pentagon officials noted that AI integration spans all technology areas, from routine tasks to kinetic operations, to ensure a dominant advantage.

Anthropic recently unveiled its Mythos AI platform for identifying software vulnerabilities, sharing it with organizations maintaining critical infrastructure but not releasing it publicly. The Pentagon distinguished this from the contract dispute, emphasizing the need to harden networks against cyber threats.

“We will become an ‘AI-first’ warfighting force across all domains." — Secretary of War Pete Hegseth The contracts reflect a broader effort under the current administration to overhaul Pentagon operations with private sector input, drawing on expertise from tech and business backgrounds.”