Can Federal Agencies Use Anthropic Claude Again? What the Blacklist Ruling Changes
Quick answer: Can federal agencies use Claude now?
Yes. On August 27, 2026, U.S. District Judge Rita F. Lin (Northern District of California) granted Anthropic summary judgment in a 59-page order, ruling that the Pentagon's "supply chain risk" blacklist of Anthropic was unlawful — and ordering the directives that barred federal agencies and defense contractors from using Claude rescinded. The Trump administration's all-agency Claude ban and Hegseth's contractor ban are vacated. For agencies and vendors serving federal or defense-adjacent clients, the legal barrier to procuring and using Claude is gone — with two caveats we cover below (§4): the government is expected to appeal, and a second, separate case in the D.C. Circuit is still pending.
The 30-second answer: can federal agencies use Claude now?
If you were holding off on Claude for government-adjacent work because of the ban, this is the "Anthropic government ban lifted 2026" answer that had been missing from search: a court struck the ban down — the government did not lift it voluntarily.
Why Anthropic was blacklisted — and the retaliation finding
The blacklist began in late February 2026, when Hegseth declared Anthropic a supply-chain risk and President Trump directed all federal agencies to "immediately cease" use of Anthropic technology within a six-month phase-out. The stated trigger: Anthropic refused to let Claude be used for fully autonomous lethal weapons or domestic mass surveillance — the company's "red lines" — and Hegseth accused the company of "arrogance and betrayal." It was the first time a U.S. company was publicly designated a supply-chain risk under the procurement statute meant to protect military systems from foreign sabotage.
Judge Lin found the government acted "based on a desire to make a public example" out of Anthropic, not on "any articulable basis" to believe the company would sabotage its models — holding the designation violated the First Amendment (unlawful retaliation), denied Anthropic Fifth Amendment due process, and was arbitrary and capricious under the Administrative Procedure Act. As the order put it: "The empty invocation of national security is not a blank check to punish and retaliate against government critics." Anthropic said the blacklist could cost it "billions of dollars in lost business and reputational harm."
What the ruling actually rescinds
- Hegseth's February 27 supply-chain-risk determination — vacated.
- The all-agency Claude ban — the Trump directive ordering federal agencies to stop using Anthropic — vacated.
- The contractor ban — Hegseth's directive barring military contractors and suppliers from doing business with Anthropic — vacated.
- Nine agencies' sanctions — including the Pentagon, Treasury, State, and DHS — removed.
Anthropic's statement: "We welcome the court's ruling that this supply chain risk designation was unlawful. We remain focused on working productively with the government."
What's still open (read this before you re-engage)
Three limits keep this from being a clean "ban fully lifted" story, and agencies should write them into their vendor risk reviews:
- The government is expected to appeal.
- A second, separately litigated designation is pending in the D.C. Circuit over a second supply-chain-risk finding that could affect civilian government contracts — that case goes to a three-judge panel, two of whom are Trump appointees who have expressed skepticism about Anthropic's arguments. Until it resolves, Anthropic "still technically remains a supply chain risk."
- The ruling does not require the Pentagon to resume working with Anthropic — it remains free to pick other vendors.
For an agency's practical posture: the blacklist itself is not a legal barrier anymore, but the D.C. case and a possible appeal are live risk flags worth monitoring before you commit a government-adjacent pipeline to Claude.
What this means for agencies evaluating Claude vs OpenAI
The ruling is a vendor-availability datapoint, not a security certification. "Is Anthropic safe for government work?" — the court found no articulable basis to believe Anthropic would sabotage its models, which removes the blacklist rationale; it does not change how you'd evaluate either vendor's FedRAMP posture, data retention, or contracting terms. Agencies that paused or switched vendors during the ban can now re-engage Claude, but the D.C. case and any appeal remain open risk flags.
If you're building an agency stack and weighing vendors, see our OpenAI-vs-Anthropic business comparison for the market-share and capability picture beyond the legal story — and keep your stack portable enough that either vendor can be swapped if a future ruling moves the other way.
Frequently asked questions
Is Anthropic banned by the US government?
No — a federal court ruled the ban illegal and vacated it on Aug 27, 2026. Anthropic technically remains under a second, separately litigated supply-chain-risk designation in the D.C. Circuit until that case concludes, and the government may appeal.
Can federal agencies use Claude?
Yes — the Aug 27, 2026 ruling vacated the Pentagon's ban on Anthropic and the directives barring agencies from using Claude, so federal agencies can use Claude again unless a pending D.C. case or a government appeal changes things.
Was the Anthropic government ban lifted in 2026?
Yes — U.S. District Judge Rita Lin lifted the 2026 government ban on Anthropic on Aug 27, ruling the Pentagon's supply-chain-risk blacklist was unlawful retaliation and ordering the designation and agency directives rescinded.
Is Anthropic safe for government work?
The court found the blacklist was "not based on any articulable basis" to believe Anthropic would sabotage its models — though a separate D.C. Circuit case over a second designation is still pending.
Can I use Claude for federal agencies now?
Federal agencies can procure and use Claude again now that the Aug 27 ruling lifted the ban, but the pending D.C. case and a possible appeal mean agencies should watch for further rulings before committing.
Weighing Claude vs OpenAI for a government-adjacent stack? Read the full market-share and capability comparison before you pick a vendor.
OpenAI vs Anthropic for Business 2026 →Or check the AI vendor-risk tracker for the latest Anthropic risk flags.
Sources
- Reuters (Aug 28, 2026) — "US judge blocks Pentagon's Anthropic blacklisting": reuters.com
- CNBC (Aug 28, 2026) — "Judge blocks Pentagon blacklist of Anthropic as supply chain risk": cnbc.com
- The Guardian (Aug 28, 2026) — "Pentagon's blacklisting of Anthropic was unlawful, US judge rules": theguardian.com
- AP (Aug 28, 2026) — "Judge says Pentagon's measures against Anthropic were 'illegal and baseless'": apnews.com
- TechCrunch (Aug 28, 2026) — "Anthropic gets its first court win over the Pentagon's supply-chain-risk label": techcrunch.com
- The Verge (Aug 28, 2026) — "Anthropic was illegally blacklisted by the Trump administration, court rules": theverge.com
- WIRED — "A Judge Has Blocked the Pentagon's Attempt to Blacklist Anthropic": wired.com
- NBC News (Aug 28, 2026) — "Federal judge blocks 'illegal' Pentagon blacklisting of Anthropic": nbcnews.com
- CNBC (Apr 8, 2026) — "Anthropic loses appeals court bid to temporarily block Pentagon blacklisting": cnbc.com
Accuracy note: This is a report on an Aug 27, 2026 summary-judgment ruling (Judge Rita F. Lin, N.D. Cal.) as covered by Reuters, CNBC, the Guardian, AP, TechCrunch, The Verge, WIRED, and NBC News between Aug 28-29, 2026. A court struck down the blacklist; the government did not lift it voluntarily. A second, separately litigated supply-chain-risk designation remains pending in the D.C. Circuit, the government may appeal, and the ruling does not require the Pentagon to resume buying Claude. The agency risk analysis is ours; it is not legal advice — consult counsel before making vendor or contract changes.