OMITTED

What the news leaves out.

← Omitted front page

Appeals Court Upholds Pentagon Blacklisting of Anthropic

7 sources · updated 2026-09-26
Left 60% Center 20% Right 20%
3 left · 1 center · 1 right

What happened

On Friday, the U.S. Court of Appeals for the D.C. Circuit ruled 2-1 to uphold the Pentagon's March designation of AI company Anthropic as a national-security supply chain risk, a label issued by Defense Secretary Pete Hegseth that bars the Department of Defense and its contractors from using Anthropic's Claude models. Judges Gregory Katsas and Neomi Rao held that the department had "ample support" for concluding that continued integration of Claude presented a statutorily covered national-security risk because Anthropic encodes restrictions that prevent the model from performing certain tasks, including uses for autonomous weapons or mass domestic surveillance that the company has refused to allow; Judge Karen LeCraft Henderson dissented. The designation followed months of negotiations over military use of Claude that ended without agreement. Anthropic challenged two distinct Pentagon designations in separate courts: a San Francisco federal judge last month found one designation unlawful and blocked related government-wide and contractor bans, while Friday's D.C. ruling upheld the other. Anthropic said it respectfully disagrees and is considering options including review by the full appeals court.
Sourcing check · Mostly on the record · 5 articles

3 of 4 articles cite a named source for the story. It was on the record from the first report. Anthropic denied part of it on the record.

What these terms mean and how they're counted
On the record · names a source 3 of 4
On the record means the article cites a source by name for the story's main claim: a person, or an institution speaking officially, such as a company statement or a named agency's spokesperson. The opposite is an unnamed source. The count is how many of the counted articles cite at least one such source; the names are everyone named across them, each once however the articles spell it. Here: US Court of Appeals in Washington, D.C., Anthropic, U.S. Court of Appeals for the District of Columbia Circuit, Pete Hegseth and U.S. Court of Appeals for the D.C. Circuit
Primary evidence · cites a document or data 4 of 4
Primary evidence means the article cites a record you could check yourself: a court filing or ruling, a dataset or report, a recording or transcript, an official statement or press release, market data. A quote is not a document. The count is how many articles cite one. Here: court ruling
Credits another outlet 1 of 4
The article says the story comes from another outlet's reporting (“first reported by Reuters”). Those articles relay the story; they are not independent confirmation. Here: POLITICO
Unnamed sources 0 of 4
The main claim is attributed to people who are not identified: “people familiar with the matter”, “officials said”. The count is how many articles do this.
On-record denial 2
A named person or institution says the claim, or part of it, is not true. It is on the record — someone signed their name to it — but it is not confirmation, so it never counts toward On the record. Criticizing or dissenting from a decision is not a denial. The count is how many people or institutions deny it; each is listed with the first article that carried the denial. Here: Anthropic (ABC News, Sep 25, 20:35 UTC); Karen LeCraft Henderson (ABC News, Sep 25, 20:35 UTC)
How it developed
When the first report landed and whether it named anyone; when the first named source and the first document appeared. Here: First report Sep 25, 16:13 UTC by New York Post, naming its source. First document in the first report, from New York Post (court ruling).
Not counted 1
Opinion, analysis and video pieces, and articles too short to assess. Every count above is out of the counted articles. An article republished under a new headline is one article.
This is about who is cited, not whether the story is true. Each source below says what it names. How we count
Omitted — what each side leaves out

Unpacked

Left-leaning coverage never reports that two federal courts split on whether the blacklist was retaliation for Anthropic's AI-safety views. CNBC's account treats the San Francisco ruling as a finding that one designation was illegal and the D.C. holding as a rejection of claims the ban was arbitrary, unauthorized, and unconstitutional. The New York Post is where Judge Rita Lin's finding of unlawful retaliation appears, along with the D.C. majority's rejection of the claim the Pentagon had "retaliated against it" for those views. The left-leaning lead in Business Insider is a grant of "sweeping authority" for the executive to tag U.S. companies as security risks. Right-leaning coverage instead opens on a victory for Trump and Defense Secretary Pete Hegseth over safety guardrails. Does Friday's ruling put the blacklist back in force, or does the San Francisco injunction still block it?
Bottom line

Read CNBC and Friday's ruling is a technical security-risk holding; only the New York Post reports that the D.C. court rejected a retaliation claim a San Francisco judge had already accepted.

The receipts — 3 facts one side's coverage lacks
  • The majority cited the prospect of overly constrained AI models shutting down unexpectedly and causing military operations to fail legal rulingin New York Post · absent from left-leaning coverage
  • Circuit Judge Karen LeCraft Henderson dissented from the 2-1 decision legal rulingin CNBC · absent from right-leaning coverage
  • Anthropic said the supply-chain-risk label would cost it hundreds of millions of dollars in revenue numberin Business Insider · absent from right-leaning coverage
The Left View
Left-leaning outlets described the ruling as a blow to Anthropic in its months-long battle with the Trump administration and as a decision that "expands the executive branch's authority to designate US businesses as national security risks." They foregrounded Anthropic's rejected arguments that the ban was arbitrary, unauthorized, and unconstitutional, and stressed that a San Francisco court had already held a parallel designation unlawful. Business Insider cited Anthropic's warning of significant "reputational stigma" and hundreds of millions of dollars in lost revenue; CNBC quoted the company remaining "confident in our position."
The Right View
Right-leaning coverage framed the decision as a victory for President Donald Trump and Defense Secretary Pete Hegseth in a dispute over safety guardrails. The New York Post emphasized that the majority found the designation reasonable after Anthropic refused those military uses, rejected Anthropic's claim of retaliation for its views on AI safety and ethics, and quoted the court's recitation of Hegseth's "deeply sobering prospect of overly constrained AI models shutting down unexpectedly and thus causing important military operations to fail." It also reported Trump and Hegseth's accusation that Anthropic was endangering American lives with "ideological posturing," which Anthropic denied, arguing AI is not yet reliable enough for autonomous weapons and that mass domestic surveillance would violate fundamental rights.
Our Take (balanced)
The left's strongest argument is that Friday's ruling, coming after a San Francisco court found a parallel designation unlawful as retaliation, confers sweeping power to blacklist U.S. firms over encoded product limits and inflicts the "reputational stigma" and large revenue losses Anthropic describes. The right's strongest argument is the majority's finding of "ample support" in restrictions Anthropic admits encoding, together with the stated risk that constrained models could fail in military operations. The central unresolved tension is whether refusing Claude for autonomous weapons or mass surveillance is a supply-chain national-security threat the Pentagon may lawfully act on, or retaliation against a company's safety stance using national-security authorities.

7 sources

The week's bottom lines, in your inbox

One email a week: the five stories that mattered and what they actually mean. Free.