Anthropic vs The Pentagon — The Story of How Anthropic Got Blacklisted
Jan 12, 2024
OpenAI Drops the Ban
OpenAI quietly removed "military and warfare" from its banned uses list. Days later, they confirmed they were already working with the Pentagon on cybersecurity tools.
Nov 7, 2024
Anthropic Goes Classified
Anthropic partnered with Palantir & AWS to deploy Claude on classified U.S. military networks — becoming the FIRST AI company to do so. But with two rules:
1. No mass surveillance of Americans.
2. No autonomous weapons.
Feb 4, 2025
Everyone drops their pledges
Google removed its 2018 promise not to build AI for weapons or surveillance. Now every major AI lab had dropped military restrictions — except Anthropic, which kept its two red lines.
Jun–Jul 2025
$800M Pentagon Contracts
The Pentagon awarded $200M contracts each to OpenAI, Anthropic, Google, and xAI for military AI. Anthropic's contract included its two restrictions. No one complained — yet.
Jan 9–12, 2026
"Won't Let You Fight Wars"
Defense Secretary Pete Hegseth released a new AI strategy demanding models be "free from usage policy constraints." At a speech he said: "We will not employ AI models that won't allow you to fight wars."
Feb 24, 2026
The 72-Hour Ultimatum
Hegseth called Anthropic CEO Dario Amodei to the Pentagon and demanded he sign a document removing ALL restrictions. Deadline: Friday 5:01 PM. The threat? Cancel the contract, blacklist the company, or force compliance by law.
Feb 26, 2026
Anthropic refuses
Amodei published a public statement refusing the Pentagon's demands. He said: "These threats do not change our position." Pentagon officials called him "a liar" with "a God-complex."
Feb 27, 2026
Trump Orders the Ban
Trump posted: "We will NEVER allow a RADICAL LEFT, WOKE company to dictate how our military fights wars." He ordered EVERY federal agency to stop using Anthropic. Hegseth then designated Anthropic a "supply chain risk" — a label previously only used for foreign adversaries like Huawei.
Feb 27, 2026
OpenAI Gets the Deal
Hours after Anthropic was blacklisted, OpenAI announced a deal to deploy on the Pentagon's classified networks. The twist? OpenAI kept the SAME two restrictions — no mass surveillance, no autonomous weapons. The Pentagon accepted their version.
Feb 27, 2026
430 Employees Fight Back
430 employees from Google and OpenAI signed an open letter called "We Will Not Be Divided" — urging their companies not to give the Pentagon what Anthropic refused. 100 Google AI engineers signed a separate internal letter.
The Real Question
Who controls AI in warfare? Point: Anthropic wasn't anti-military — they were the MOST embedded AI company in defense. The fight was about two narrow lines: no spying on Americans, no killer robots without human control. OpenAI got the deal with the exact same restrictions. Was this about policy — or politics?
Follow
@quitehacker for more!
#anthropi #openai #ClaudeAI #chatgpt #trump #DonaldTrump