The legal fight between Anthropic and the Pentagon over the company's blacklisting as a "supply chain risk" went to an appeals court on Tuesday. The judges appear divided. But the story is no longer just about Anthropic and the Department of Defense. It is about how much of Silicon Valley is willing to publicly line up against the Trump administration.

The answer, as of this week: a lot of it.

19 OpenAI Researchers Backed Anthropic. Read That Again.

An amicus brief filed with the court on Monday carries the names of 37 top AI researchers. Nineteen of them work at OpenAI. Ten work at Google DeepMind. The group also includes Google chief scientist Jeff Dean, who signed in his personal capacity.

Think about what that means. OpenAI just signed a Pentagon contract days after Anthropic was ejected from the vendor list. OpenAI is Anthropic's direct competitor. And yet 19 of its own researchers filed a court document saying the Pentagon was wrong to blacklist Anthropic for maintaining safety guardrails. That is not corporate posturing. That is researchers at competing companies agreeing that the government's actions threaten the entire field.

"We wanted to make sure we were arming the court with an understanding of the industry's perspective," said Nicole Schniedman, a Protect Democracy attorney whose name appears on the brief. "It is unprecedented to label a domestic company a supply chain risk for taking a stand on safety guardrails."

The Core Dispute: Safety Rules vs. Military Contracts

The backstory is worth understanding. The Pentagon (which now officially calls itself the Department of War) was angered by Anthropic's refusal to drop two specific policies: one prohibiting the use of Claude for targeting autonomous weapons, and another prohibiting the synthesis of data from mass surveillance of U.S. citizens.

In response, Defense Secretary Pete Hegseth designated Anthropic a "supply chain risk," a classification normally reserved for companies in adversary nations like China and Russia. The practical effect: Anthropic can no longer do business with the government or any government contractor. The company says it could lose hundreds of millions of dollars.

Anthropic's lawsuit calls the designation "unprecedented and unlawful" and alleges retaliation for exercising First Amendment rights. Neither of Anthropic's two lawsuits seeks to force the government to contract with the company. They want the blacklisting label removed.

The Cloud Giants Are Not Backing Down Either

Microsoft filed a separate amicus brief in support of Anthropic, urging the court to grant a temporary restraining order delaying the blacklisting while the case is heard. All three major cloud providers (Microsoft Azure, Google Cloud, Amazon AWS) have publicly stated they will continue distributing Anthropic's Claude models through their platforms, though not for defense-related work.

This is the part that should make the Trump administration nervous. Many of the tech industry's biggest names (Marc Andreessen, David Sacks, Elon Musk, Sundar Pichai, Tim Cook, Jensen Huang) supported Trump's 2024 campaign with the expectation of minimal government oversight during the AI buildout. The implicit deal was: stay out of our way, and we will stay on your side.

The administration may have calculated that going after Anthropic was safe because CEO Dario Amodei did not fund Trump's campaign or attend his inauguration. Other companies, OpenAI and xAI and Google, were ready to fill the gap. But the calculation appears to have been wrong. What started as a narrow contract dispute has escalated into something the administration cannot easily contain: the entire AI research community, including employees at companies that directly benefit from Anthropic's exclusion, publicly challenging the government's authority.

What Happens Next

The appeals court judges appeared divided during oral arguments. The Pentagon has not provided a clear legal justification for the blacklisting. Hegseth's original announcement on X made no legal argument. Trump's Truth Social post told agencies to "cease all use of Anthropic's technology" but did not invoke supply-chain-risk authority. The formal letter to Anthropic contained no legal reasoning either.

Anthropic is not asking the court to force the government to buy Claude. It is asking for the right to not be labeled a national security threat for refusing to build autonomous weapons targeting systems. The fact that this is even a question tells you everything you need to know about where the AI industry's relationship with the U.S. government stands in 2026.

First reported by Associated Press. Additional reporting from Federal News Network, Fast Company via DeviceDaily, and CNBC.