PENTAGON STRIKES AI DEALS WITH SEVEN TECH GIANTS AFTER SHUTTING OUT ANTHROPIC 

THE PRESS CENTER | THE MORSE CODE | TECHNOLOGY

MAY 2 2026 AT 4:57 PM

The Department of Defense on Friday unveiled agreements with seven major technology companies to deploy their artificial intelligence systems across classified military networks—formalizing one of the largest AI procurements in U.S. defense history.

Conspicuously absent is Anthropic. The Trump administration previously barred the company from federal contracts after Anthropic refused to allow the Pentagon to use its Claude models for “all lawful purposes,” a phrase the government interprets to include autonomous weapons and expansive surveillance. The administration labeled Anthropic a “supply chain risk,” a designation historically reserved for firms tied to foreign adversaries. A federal judge in California blocked that designation last month after Anthropic sued.

The new Pentagon agreements include AI systems from SpaceX, OpenAI, Google, Microsoft, Nvidia, Amazon Web Services, and Reflection. Several of these companies—such as Palantir and OpenAI—already hold defense AI contracts, but the breadth of Friday’s announcement signals a major expansion of the military’s AI footprint.

The Pentagon said the tools will support “lawful operational use” and accelerate its shift toward an “AI‑first fighting force,” enabling U.S. personnel to maintain “decision superiority across all domains of warfare.” Officials also highlighted rapid adoption of the department’s GenAI.mil platform, which they say has been used by 1.3 million DoD employees.

Anthropic’s exclusion gives the administration leverage as it reopens talks with the company. The One Big Beautiful Bill Act set aside substantial funding for AI and offensive cyber operations, and the companies selected Friday now have access to that revenue stream—while Anthropic does not.

Still, discussions have resumed. Anthropic CEO Dario Amodei met last month with White House Chief of Staff Susie Wiles following the company’s announcement of Mythos, a tool capable of identifying cybersecurity threats and, controversially, mapping potential attack vectors.

Until recently, Anthropic’s Claude was the only AI model operating inside the Pentagon’s classified environment. That era appears to be over.

CNN’s Hadas Gold and Sean Lyngaas contributed reporting.

SOURCE CREDIT:CNN.COM