On August 27, a federal judge issued a 59-page ruling voiding the Pentagon's designation of Anthropic as a national security "supply chain risk," finding the government's case was not just wrong but unconstitutional. It's a genuine, substantive win for Anthropic, its most significant so far in a legal fight that started in March. It's also not the end of the story. A separate designation, under a different statute, is still in effect, and Anthropic remains blocked from Department of War contracts specifically while that second track continues in a different court.

59
Pages in Judge Rita Lin's summary judgment order voiding the designation
2
Separate legal tracks, under two different statutes, still running in two different courts
6
Months since the designation was first announced in March

How this started

In February, the government began pressuring Anthropic to drop two positions it had held in contract negotiations: refusing to let Claude be used for mass surveillance of Americans, and refusing to permit its technology to power fully autonomous weapons without human oversight over targeting and firing decisions. Anthropic held its position. Defense Secretary Pete Hegseth responded on social media, and in March the Pentagon formally designated Anthropic a supply chain risk, a label that had previously been reserved for companies tied to foreign adversaries. The designation barred Anthropic from Department of War contracts, required contractors including Amazon, Microsoft, and Palantir to certify they weren't using Claude in Pentagon-related work, and came alongside a separate presidential directive ordering all federal employees, not just the military, to stop using Claude entirely.

Anthropic sued twice, once in the Northern District of California and once in the D.C. Circuit Court of Appeals, arguing the designation violated the Administrative Procedure Act, the First Amendment, and the Fifth Amendment. Dozens of researchers at rival labs OpenAI and Google DeepMind filed an amicus brief in their personal capacities supporting Anthropic, an unusual moment of cross-competitor solidarity, arguing the designation could damage U.S. AI competitiveness and chill honest public discussion of AI safety risks industry-wide.

What the August 27 ruling actually found

Judge Rita Lin, N.D. Cal.Lin's order found the Pentagon's actions "constituted unlawful retaliation in violation of the First Amendment" and that Anthropic "was denied the pre-deprivation process required under the Fifth Amendment." She wrote that while the military is owed real deference on national security matters, its case against Anthropic wasn't built on one: "the evidence demonstrates that the broad measures imposed on Anthropic were illegal and baseless." On the government's argument that Anthropic's public criticism meant it couldn't be trusted, Lin wrote plainly that "neither the Constitution nor the federal statute invoked by Defendants allows them to impose sweeping penalties based principally on Anthropic's critique of the Administration's views." She also singled out the government's public statements, including remarks about Anthropic's "arrogance" for raising concerns about how its products would be used, as evidence the designation was about punishing a critic rather than addressing an actual security risk.

An Anthropic spokesperson said the company welcomed the ruling. Industry trade group ITI's president called it a step that "removes uncertainty for businesses across industry seeking to partner with the U.S. government," a framing that reaches well beyond Anthropic to any company that might someday find itself on the wrong side of a public disagreement with a federal customer.

Why Anthropic is still partly blocked

The August 27 ruling resolves one of the two designations, the one made under 10 U.S.C. § 3252 and the presidential directive. A second designation, made separately under a different statute, 41 U.S.C. § 4713, is being litigated on its own track in the D.C. Circuit, and it remains in effect. Back in April, that court declined to pause enforcement of the § 4713 designation while the case proceeded, finding the balance of harms favored the government during "an active military conflict," and set an expedited schedule that included oral argument on May 19. At that argument, the judges appeared genuinely split, with one, Judge Karen LeCraft Henderson, stating she saw "no evidence to support the Pentagon's determination that Anthropic poses a supply-chain risk." No final ruling on that track has been reported as of this writing.

Practically, that means Anthropic can work with commercial clients and non-Pentagon federal agencies without restriction, but remains excluded from Department of War-specific contracts under the surviving designation. The government is expected to appeal the California ruling as well, so even the track Anthropic just won isn't necessarily finished.

The bottom line: "Illegal and baseless" is a real, hard-won characterization from a federal judge, not a technicality, and it validates what Anthropic argued from the start: that a company's public disagreement with a government customer isn't a national security threat. But it settles one of two separate legal fights, not the whole dispute. A vendor watching this case for a signal about how the government treats contractors who push back publicly should read both tracks, not just the one that went Anthropic's way.

Sources: AI Pulse · Compliance Watch · workplaceai.ai. The February-March origins of the dispute, the two lawsuits, and the OpenAI/DeepMind amicus brief: TechCrunch and PBS NewsHour (AP), March 9, 2026; Lawfare, March 9, 2026. Judge Lin's August 27 summary judgment ruling and its specific findings and quotes: CNBC, August 28, 2026; CNN via ABC17News, August 27, 2026; Washington Technology, August 2026; Nextgov/FCW, August 2026; TFTC, September 2026. The D.C. Circuit's April stay denial, the "active military conflict" language, and the May 19 oral argument: Reason/Volokh Conspiracy, April 8, 2026; Federal News Network, May 20, 2026; Kilpatrick Townsend legal alert, April 9, 2026. The current status of both tracks: the independent Anthropic v. Department of War case tracker at anthropic-v-dow.vercel.app. Every quote and figure above is attributed to its original reporting; none is a WorkplaceAI study.