Anthropic’s Pentagon Fight Gets a Rule-of-Law Boost
Wednesday, June 24, 2026 · 12 min
Anthropic Pentagon litigation features a rule-of-law amicus attack on the supply-chain-risk label, while background reports show how Claude’s weapons-and-surveillance red lines collided with DoD demands for “all lawful purposes” access.
Listen
Show notes
Anthropic Pentagon litigation features a rule-of-law amicus attack on the supply-chain-risk label, while background reports show how Claude’s weapons-and-surveillance red lines collided with DoD demands for “all lawful purposes” access.
In this episode
- Amicus Brief in Anthropic PBC vs. Department of War - Society for the Rule of Law — Society for the Rule of Law
 # Amicus Brief in Anthropic PBC vs. Department of War Several charter members of the **Society for the Rule of Law—**including our Board President **Alan Raul**—have joined in filing an amicus…
- Step Back — Step back for me: when an AI company says it will support national-security work but not weapons or surveillance, how are those red lines actually written into Pentagon contracts and enforced in practice — is there a technical off-switch, audits, penalties, or mostly trust?
Background sources
- Pentagon vs. Anthropic: Autonomous Weapons AI Guardrails and the Governance Crisis for Enterprise AI Vendors – Lab Space — Cloudsecurityalliance
- OpenAI says it shares Anthropic's 'red lines' over military AI use - OPB — Opb
- Deadline looms as Anthropic rejects Pentagon demands it remove AI safeguards | HPPR — HPPR
- Deadline looms as Anthropic rejects Pentagon demands it remove AI safeguards — Knba
- OpenAI Claims Safety 'Red Lines' in Pentagon Deal—But Users Aren't Buying It - Decrypt — Decrypt / Jose Antonio Lanz
- OpenAI to amend DoD AI contract after surveillance and autonomous weapons concerns - DCD — Datacenterdynamics
- Anthropic AI rejects Pentagon's weapons & surveillance ultimatum — CNN
# Anthropic AI rejects Pentagon's weapons & surveillance ultimatum ## CNN 19600000 subscribers 2853 likes ### Description 296722 views Posted: 27 Feb 2026 American artificial intelligence company Anthropic could be at risk being designated a “supply chain risk” — a label typically reserved for companies tied to foreign adversaries. The Pentagon, which uses Anthropic’s Claude AI system on its…
- OpenAI Vs. Anthropic: How the Pentagon Picked Its Partner — The New York Times
# OpenAI Vs. Anthropic: How the Pentagon Picked Its Partner ## Hard Fork 588 likes ### Description 26331 views Posted: 1 Mar 2026 On Friday, President Trump ordered federal agencies to stop using Anthropic’s A.I. systems and Defense Secretary Pete Hegseth designated the company a “supply chain risk.” Then, just a few hours later, the OpenAI chief executive, Sam Altman, announced that his…
- DoD Directive 3000.09, "Autonomy in Weapon Systems," January 25, 2023
D OD D IRECTIVE 3000.09 # A UTONOMY IN W EAPON S YSTEMS Originating Component: Office of the Under Secretary of Defense for Policy Effective: January 25, 2023 Releasability: Cleared for public release. Available on the Directives Division Website at https://www.esd.whs.mil/DD/. Reissues and Cancels: DoD Directive 3000.09, “Autonomy in Weapon Systems,” November 21, 2012 Approved by:…
“For what it’s worth, Program Offices and contracting officers are not including this (or any limitations on autonomous control) as a blanket or even specific requirement to be in compliance for any contract awards, so it’s largely irrelevant. There’s also no independent V&V team…” — Hacker News (9 pts thread)
Our take: We think this gets at the whole gap between policy theater and procurement reality: a directive is only as strong as the contract language, testing regime, and people paid to say no. If there’s no independent verification, “human judgment” becomes a slogan with a PDF number.
“>1.2. POLICY. >a. Autonomous and semi-autonomous weapon systems will be designed to allow >commanders and operators to exercise appropriate levels of human judgment over >the use of >force. Should state: will be designed to mandate and require . . . "allow" is obviously inserted…” — Hacker News (1 pts thread)
Our take: That close read may sound lawyerly, but “allow” versus “mandate” is exactly the kind of verb choice that matters when autonomy meets lethal force. Wiggle room is not a bug in these documents; often, it is the point.