# Anthropic v. Pentagon Supply Chain Risk Designation

**Type:** Legal case / Governance precedent  
**Status:** Active litigation (May 19, 2026 oral arguments scheduled)  
**Jurisdiction:** DC Circuit Court of Appeals  
**Significance:** First judicial test of constitutional protection for voluntary AI safety constraints excluding military applications

## Overview

Legal challenge to Secretary of Defense Hegseth's March 3, 2026 designation of Anthropic as a "supply chain risk" under 10 U.S.C. § 2339a, requiring defense contractors to certify they do not use Claude. The designation was based on Anthropic's refusal to allow Claude for fully autonomous lethal weapons or mass surveillance of Americans.

## Case Chronology

- **2026-03-03** — Secretary of Defense Hegseth designated Anthropic as supply chain risk under 10 U.S.C. § 2339a, requiring defense contractors to certify non-use of Claude. Stated basis: Anthropic's refusal to allow Claude for fully autonomous lethal weapons or mass surveillance.
- **2026-03-09** — Anthropic filed suit in DC district court challenging designation
- **2026-03-26** — DC district court granted preliminary injunction blocking designation, characterizing it as "classic illegal First Amendment retaliation" (citing parallel California N.D. ruling)
- **2026-04-08** — DC Circuit reversed stay in 3-judge panel ruling. Court framed harm as "primarily financial in nature" and applied equitable balance test, finding government's wartime AI procurement management interest prevailed over Anthropic's financial harm
- **2026-05-19** — Oral arguments scheduled. Court directed briefing on threshold jurisdictional questions including whether DC Circuit has jurisdiction over the petition

## Current Legal Posture

**DOD ban:** STANDING (DC Circuit denied stay)  
**Other federal agency ban:** BLOCKED (N.D. California injunction)  
**Merits:** Not yet decided

## Judicial Framing Divergence

**N.D. California:** Pentagon action constitutes First Amendment retaliation. Constitutional harm requiring protection.  
**DC Circuit:** Anthropic's harm is "primarily financial." Administrative law equitable balance test applies, not constitutional scrutiny.

This framing divergence determines whether voluntary corporate safety constraints have constitutional protection or can be administratively dismantled.

## Governance Implications

If DC Circuit's financial framing prevails:
- Voluntary AI safety constraints excluding military applications have no constitutional floor
- Can be administratively dismantled through supply chain risk designation
- Every AI lab with safety policies excluding certain military uses faces same designation risk
- Creates jurisdictional boundary: civil commercial jurisdiction protects voluntary constraints, military procurement jurisdiction does not

## Related

- [[voluntary-ai-safety-constraints-lack-legal-enforcement-mechanism-when-primary-customer-demands-safety-unconstrained-alternatives]]
- [[three-track-corporate-safety-governance-stack-reveals-sequential-ceiling-architecture]]
- [[strategic-interest-alignment-determines-whether-national-security-framing-enables-or-undermines-mandatory-governance]]