The Anthropic–Pentagon split

As of 19 October the transition is treated as done by the supplier side and is being contested in court by Anthropic. Palantir's chief technology officer said on television that there has been no operational impact from the Pentagon dropping Anthropic and that Claude is already replaced, naming no replacement for any role. On 13 October Anthropic filed a petition for panel rehearing or rehearing en banc in the D.C. Circuit in Anthropic PBC v. U.S. Department of War. CNBC quoted the filing: "The Secretary of War branded Anthropic — an American AI company whose models help defend the nation — a national-security 'supply chain risk' because it declined to permit its models to be used for mass surveillance of Americans or lethal autonomous warfare and publicly explained those AI-safety commitments. A divided panel upheld that extraordinary designation." Anthropic argues the court ignored "straightforward evidence the designation was retaliation for Anthropic's speech about AI safety."

The filing sets up a conflict between two courts. Reporting on the petition says Anthropic contrasts the panel's First Amendment holding with an August ruling by a San Francisco federal judge, who found that the Department of Defense had violated Anthropic's First Amendment rights and that the supply-chain-risk designation amounted to illegal retaliation, and calls the two irreconcilable. The petition also argues that the majority's reading of "otherwise manipulate" conflicts with Fischer v. United States, and quotes Judge Henderson's dissent: "Agree to the Secretary's demands or risk being designated a national security threat."

The second half of the arc's question remains unanswered. No department document has been published saying what replaced Claude in which role, and none says what replaces the contract safeguards Anthropic refused to relax. A Pentagon spokesperson told CNBC the agency does not comment on pending litigation, and Anthropic declined to comment beyond the filing. The court has not ruled on the petition.

As of 12 October the department has said in words what no document has yet recorded. A Department of Defense official said in a statement in the first week of October that the Pentagon "has ceased the use of Anthropic products", the BBC reported — the first time the department has said the transition is done. The BBC adds that Defence Secretary Pete Hegseth labelled Anthropic a supply chain risk in February, set a six-month phase-out at the end of that month and announced the Pentagon would stop using it by late August, and that "It is not clear why there has been a delay." No filing, memorandum or published department document records the completion, and the statement names no replacement. The BBC is the only outlet carrying it.

What the record does contain is the other side's account. On CBS's "Face the Nation", Palantir chief technology officer Shyam Sankar said "there's been no operational impact from that", that the Department "has a wide range of model providers, everything from XAI and OpenAI to open-weight models like Nemotron from NVIDIA", and "It's already been replaced. It's already been replaced." He named no provider for any specific role and gave no figures for the volume of work moved, and the Department has not confirmed his assessment. Multiple people familiar with the matter told the BBC that Claude was still being used by the department in research, analysis and intelligence gathering as well as in military operations against Iran as recently as the week before the statement, and that the Pentagon has signed contracts with Google, xAI and OpenAI.

On the safeguards this question also tracks, the supplier moved and the government did not. Anthropic published a Usage Policy update, effective November 12, stating that "The updated section makes clear that our prohibitions include the software and components that make weapons work, as well as actions like arming drones and other autonomous vehicles", and that "Claude cannot be used to decide or recommend who to investigate, arrest, or charge in a law enforcement or criminal justice process. We also prohibit Claude from being used to build or improve tools designed for surveillance." Anthropic says both changes reflect how it has enforced its previous policy. The post does not mention the Pentagon, any government contract or the designation, and nothing published in this period says what, if anything, replaces the contractual prohibitions the department refused to relax.

As of 28 September the split has a published court judgment and the Department's reasons are on the record in the government's own words. The D.C. Circuit decided Anthropic PBC v. United States Department of War in late September. Circuit Judge Katsas wrote for the court, Circuit Judge Henderson dissented, and CNBC and Courthouse News both report the vote as 2-1 with Circuit Judge Rao joining. The opinion states that the Department excluded Claude from its supply chain under the Federal Acquisition Supply Chain Security Act after Anthropic refused to relax contractual prohibitions on the use of Claude for lethal autonomous warfare or domestic surveillance, and that the Department had ample support for its conclusion that continued integration of Claude presented a statutorily covered national-security risk. It records that Under Secretary for Research and Engineering Emil Michael cited Anthropic's refusal to allow all lawful uses and its ability to alter system guardrails and model weights. Anthropic's due-process and First Amendment claims were rejected.

What the ruling does not settle is the thing this storyline tracks. The opinion never mentions GenAI.mil, the Maven Smart System, or any model that replaced Claude; the transition status rests where it did, on one official's word. Henderson's dissent argues the case turns entirely on the scope of the residual clause and warns that a future contractor faces a choice between agreeing to the Secretary's demands or being designated a national security threat. An Anthropic spokesperson said the company is considering all options, including further review, and noted that another federal court has held a parallel designation unlawful; Breaking Defense reports the remaining routes are a discretionary en banc rehearing or a Supreme Court petition. No department statement or filing confirming completion of the transition has been published.

As of 14 September the transition is nearly done, on one official's word. Emil Michael, Under Secretary of Defense for Research and Engineering, told DefenseScoop that about 90% of classified AI workloads have moved off Anthropic, with the rest due by the end of September; DefenseScoop reports OpenAI's ChatGPT, xAI's Grok and Google's Gemini are being deployed in its place, and that the break followed Anthropic's attempt to secure contract terms preventing use for mass surveillance of US citizens or fully autonomous lethal weapons, which the department rejected. No other outlet had matched the figure as of this snapshot, and no contract documents have been published.