A federal judge just released Trump's own Truth Social post as evidence against his Pentagon

Federal judge rules Pentagon Anthropic blacklist unconstitutional retaliation 2026

The Ruling

Anthropic v. Department of War

59 pages

Length of Judge Rita Lin's written summary judgment order, filed August 27

First ever

A domestic US company designated a "supply chain risk," per the court's own record

$180M+

In deals the court record shows collapsed within days of the blacklist

2 lawsuits

Anthropic's parallel case in the D.C. Circuit remains pending

Published: August 29, 2026 | Updated: August 29, 2026 | Category: AI | By Mahesh

On the night of August 27, U.S. District Judge Rita F. Lin issued a 59-page order permanently barring the Department of Defense from enforcing its designation of Anthropic as a national security "supply chain risk," ruling that the designation and the broader campaign against the company were unconstitutional retaliation for Anthropic's public criticism of the Pentagon's AI policy, according to multiple outlets that reviewed the ruling, including Axios and Al Jazeera.^[1] "The empty invocation of national security is not a blank check to punish and retaliate against government critics," Lin wrote. That final summary judgment order builds on and confirms an earlier ruling in the same case, filed directly on the public court docket and available in full through Justia's case law database, docket number 3:26-cv-01996 in the U.S. District Court for the Northern District of California.^[2] That earlier order, granting Anthropic a preliminary injunction back in March, laid out in granular detail exactly what the government did, and it is worth reading directly rather than through secondhand summary.

What the Government Actually Said, In Its Own Words

The court's own record reproduces the specific statements that triggered this case, and they are more direct than most coverage has conveyed. On February 27, 2026, at 3:47 p.m. Eastern, President Trump posted on Truth Social that he was "directing EVERY Federal Agency in the United States Government to IMMEDIATELY CEASE all use of Anthropic's technology," calling the company a "RADICAL LEFT, WOKE COMPANY" and its position a "DISASTROUS MISTAKE trying to STRONG-ARM the Department of War." Just over an hour later, Defense Secretary Pete Hegseth posted on X that Anthropic had delivered "a master class in arrogance and betrayal," accused the company of "Silicon Valley ideology," and directed the Department of War to formally designate Anthropic a supply-chain risk, stating the decision was "final." Both statements are quoted verbatim and in full in the court's own order, which treated the language itself, not just the resulting policy, as central evidence.

The dispute behind those statements traces to contract negotiations that began in fall 2025 over deploying Claude on the Pentagon's "GenAI.mil" platform. According to the court record, Anthropic agreed to let the Department of War use Claude for essentially all lawful purposes, with two exceptions it declined to remove: mass surveillance of Americans and lethal autonomous weapons. Anthropic's own head of policy testified, without the government disputing it, that these negotiations remained "cordial and amicable" through late February. The government's justification for the blacklist rested on Anthropic's supposed technical ability to sabotage deployed systems, but the court noted that Anthropic submitted unrebutted evidence that it has no such access once its models are running inside government-secured systems, and that, at oral argument, the government's own counsel admitted he was unaware of any evidence supporting the sabotage theory.

"Punishing Anthropic for bringing public scrutiny to the government's contracting position is classic illegal First Amendment retaliation." — Judge Rita F. Lin, Order Granting Motion for Preliminary Injunction, filed March 26, 2026, in the case's public court record^[2]

The Timeline the Record Reveals

Reading the court's chronology closely surfaces a detail that undercuts the government's own stated justification. According to the order, Secretary Hegseth formally designated Anthropic a supply-chain risk on March 3, based on a memo from Under Secretary Emil Michael describing the company as an "unacceptable national security threat." Yet the court's record includes an email exchange showing that on March 4, the day after that designation was finalized, Michael and Anthropic CEO Dario Amodei were still cordially trading draft contract language, with Michael writing to Amodei: "After reviewing with our attorneys and seeing your last draft (thanks for being fast), I think we are very close here." Judge Lin's order calls it "exceedingly difficult to square this correspondence" with the government's contemporaneous internal characterization of Anthropic as a hostile, untrustworthy actor, and treats the contradiction as direct evidence that the stated national-security rationale was pretextual.

The court's record also documents the immediate commercial fallout Anthropic experienced. Within days of the blacklist, the General Services Administration removed Anthropic from its federal AI platform, the Treasury Department and Federal Housing Finance Agency announced they were terminating Claude use entirely, and Lawrence Livermore National Laboratory shut down its access. Anthropic's own sworn filings, cited directly in the order, stated that three deals worth more than $180 million collapsed within days despite being close to closing, and that the company's chief financial officer projected Anthropic could lose "between hundreds of millions and multiple billions of dollars" in 2026 revenue depending on how broadly customers interpreted the blacklist.

Why the Court Rejected the Government's Central Argument

The legal core of the ruling turns on a narrow but consequential statutory question: what "supply chain risk" actually means under the specific law the Pentagon invoked, 10 U.S.C. § 3252. The court's order traces that statute back to its 2011 legislative origin, when Congress was responding to concerns about "attacks on its systems and networks" from "foreign intelligence, terrorists or other hostile elements," according to the Senate Armed Services Committee report the court cites directly. Judge Lin's order states plainly that the statute is "directed at covert acts or hacks, not overt positions taken during contract negotiations," and that the government's theory, that a vendor becomes an "adversary" simply by publicly pushing back on contract terms, is "deeply troubling and inconsistent with the statutory text."

The court also found the government failed the law's own required procedure, separate from the constitutional violation. Section 3252 requires the Secretary of Defense to make a written determination that "less intrusive measures are not reasonably available" before issuing a supply-chain designation, and to explain that reasoning to Congress. The court's order states that nothing in the government's own submitted record shows this analysis was actually performed, and that Hegseth's letters to congressional committees contained no such discussion at all, a gap the government's own counsel conceded at oral argument.

What This Means for Anthropic's Path to an IPO

This ruling lands at a specific moment for Anthropic that gives it weight well beyond the underlying legal dispute. Depth Grid covered earlier this week how Anthropic's confidential IPO filing is expected to list broader AI backlash as a formal risk factor, and this specific legal exposure, a government blacklist that had already cost the company hundreds of millions in collapsed deals, was very likely a distinct line item investors were pressing the company's leadership about directly. CNBC's reporting on the ruling noted it "clears a significant hurdle" for Anthropic as it moves toward what is expected to be a near-record public listing, since resolving this specific overhang removes one of the more quantifiable, near-term risks a prospective public shareholder would have had to price into the stock.

The victory is not complete, and the court record itself is explicit about that. Anthropic has a second, parallel lawsuit challenging a related Pentagon designation currently pending before a three-judge panel of the D.C. Circuit Court of Appeals, a case governed by a different statute than the one Judge Lin ruled on. Multiple outlets, including the Associated Press, reported that a separate federal appeals panel in Washington had, in an earlier and distinct proceeding, declined to block enforcement while that case continues, a reminder that this dispute is being fought on two separate legal tracks simultaneously and that Thursday's ruling, however sweeping in its language, resolves only one of them.

What to Watch Next

The Justice Department is expected to appeal Judge Lin's ruling, according to NPR's reporting on the decision, meaning the Ninth Circuit Court of Appeals is the next venue where this specific designation will be tested. The outcome of the still-pending D.C. Circuit case, governed by a separate statute covering exclusion from civilian government contracts rather than Defense Department systems specifically, remains the other open thread. Until both tracks resolve, Anthropic's relationship with the federal government as a customer remains only partially restored, even though this week's ruling removes the most sweeping and most publicly damaging element of the dispute.