Developing Story
Pentagon Claude Ban Litigation — Conflicting Circuit Court Decisions
The D.C. Circuit Court of Appeals upheld the Pentagon's ban on Anthropic's Claude on September 25, 2026, citing national security risk under 41 U.S.C. § 4713 — conflicting with an earlier ruling (reportedly from the Northern District of California) that found the ban illegal under the supply-chain risk statute 10 U.S.C. § 3252. The split creates legal uncertainty over federal agencies' authority to blacklist specific AI vendors and may require further appellate review.
This page was written and is updated by the paper from the source articles listed below. Importance is how central the topic is to current coverage; confidence is how well-sourced the page is; mentions counts the articles that fed it. How the wiki works →
## Overview
The U.S. Defense Department's ban on Anthropic's Claude AI model has become the subject of conflicting federal court rulings, creating legal uncertainty over the Pentagon's authority to restrict specific AI vendors. The split traces to two different statutory bases: a "supply chain risk" designation under 10 U.S.C. § 3252 and a national security risk classification under 41 U.S.C. § 4713.
## Timeline
- A federal court (reportedly the Northern District of California) ruled approximately a month prior to September 25, 2026 that the Pentagon's ban on Claude is illegal, reportedly on "supply chain risk" grounds under 10 U.S.C. § 3252 (SiliconANGLE, September 25; ChatGPT Is Eating The World, September 25).
- On September 25, 2026, the D.C. Circuit Court of Appeals upheld the Pentagon's ban on Claude, siding with the Department of War's classification of the Anthropic model as a national security risk under 41 U.S.C. § 4713 — directly conflicting with the earlier ruling (SiliconANGLE, September 25; ChatGPT Is Eating The World, September 25).
- The diverging decisions reportedly stem from the fact that the matter is governed by two different pieces of legislation (SiliconANGLE, September 25).
## Why It Matters
This case sits at the intersection of federal procurement law, national security policy, and the rapidly evolving legal landscape around government use of AI systems. The split between courts — one applying supply-chain risk statute 10 U.S.C. § 3252, the other national security risk statute 41 U.S.C. § 4713 — suggests the issue may require further appellate or Supreme Court review to resolve. The outcome will shape whether federal agencies can unilaterally blacklist specific AI vendors and under what statutory authority such bans can be challenged. Notably, one source refers to the Pentagon as the "Department of War," suggesting a possible renaming or rebranding not yet independently confirmed.
## Key Entities
- **Anthropic / Claude**: The AI model and its developer at the center of the ban.
- **U.S. Defense Department (Pentagon) / "Department of War"**: The agency that imposed the ban; referred to as "Department of War" in one September 25 source (ChatGPT Is Eating The World, September 25) — naming unconfirmed elsewhere.
- **D.C. Circuit Court of Appeals**: Issued the September 25 ruling upholding the ban under 41 U.S.C. § 4713.
- **Northern District of California (reported)**: Court reportedly behind the earlier ruling finding the ban illegal under 10 U.S.C. § 3252 (ChatGPT Is Eating The World, September 25).
## Open Questions
- Which statute will ultimately govern the outcome given the two rulings rely on different legislation (10 U.S.C. § 3252 vs. 41 U.S.C. § 4713)?
- Will this be appealed further, potentially to the Supreme Court?
- How does this interact with broader Pentagon AI vendor blacklisting efforts (e.g., DJI)?
- Is "Department of War" an official renaming of the Defense Department, or informal/alternate terminology used by one source?