◉ PSYCHOHISTORY

Hegseth Ultimatum (Feb 27 2026)

eventAI & Compute · Nations & Geopolitics · Defense & Military-Industrial
The Pentagon set a Friday deadline for an AI company to open its models to the military, and the company said no.
Who they are

The February 27, 2026 ultimatum from US Defense Secretary Pete Hegseth to the AI company Anthropic[1].

What they do

The engine logs it as a documented showdown proving 1950-era emergency powers are still live and aimed at AI companies.

How it works

Hegseth demanded Anthropic open Claude for unrestricted military use by 5:01 PM that Friday or lose a $200 million classified contract; CEO Dario Amodei publicly refused, saying 'we cannot in good conscience accede to their request,' the administration threatened to invoke Defense Production Act Title III (from 1950), and Anthropic was labeled a national-security supply-chain risk (with a contradictory 6-month grace period). OpenAI stepped in to provide models on classified networks, and the WSJ later confirmed the Defense Department had been using Claude via Palantir for Iran strikes and Venezuela operations.

Why it matters

It matters because it shows early-Cold-War legal authorities from 1950 are still active and that leaders were ready to use them against frontier-AI companies, the modern equivalent of once wielding those powers against industrial giants.

What's new (Aug–Oct 2026)

Aug 16, 2026 — The New York Times reported that in mid-July the Air Force warned many of the biggest military contractors that by Sept. 1 all weapons and control-system software must be free of Anthropic products; within a month came "an unexpected reversal". (nytimes.com)
Aug 27, 2026 — Federal judge Rita Lin ruled on Thursday evening, in a 59-page decision, that labeling Anthropic a supply-chain risk was illegal. (theguardian.com ; techcrunch.com)
Aug 31, 2026 — The Pentagon launched its own versions of ChatGPT and Grok for 3 million civilian and military personnel (TechCrunch). Its GenAI.mil system counts 1.7 million users among that 3 million workforce (Fortune). (techcrunch.com ; fortune.com)

The engine's record — word for word
Formal documented ultimatum from US SecDef Pete Hegseth to Anthropic: open Claude architecture for unrestricted military use by Fri Feb 27 at 5:01 PM or face immediate termination of a $200M classified defense contract (awarded July 2025). Anthropic CEO Dario Amodei publicly refused: 'we cannot in good conscience accede to their request.' Trump administration threatened to invoke Defense Production Act Title III (1950). Hegseth simultaneously designated Anthropic a 'Supply-Chain Risk to National Security' — with a six-month grace period (contradiction indicating operational dependency). OpenAI immediately stepped in to provide models on classified networks. WSJ subsequently confirmed DOD had been using Claude via Palantir for Iran kinetic strikes and Venezuela operations (intel assessments, target ID, battle simulation). **Report #87:** the Feb 27 2026 ultimatum to Anthropic + DPA Title III invocation threat operates at the same statutory layer as the original Defense Production Act of 1950 — proving the 1950-era authorities remain active and Trump-2-era leadership prepared to invoke them against frontier-AI companies. The ultimatum is the kinetic enforcement equivalent at the algorithmic-warfare substrate layer of the early-Cold-War invocation of these authorities against industrial primes. [Live pass Oct 6 2026] 2026-08-16 (NYT publication) — NYT: In mid-July, the Air Force warned many of the biggest military contractors that by Sept. 1 all weapons and control systems software must be free of Anthropic products; within a month, "an unexpected reversal". (1 source) (nytimes.com) | ~2026-08-27 — Federal judge Rita Lin ruled (Thursday evening) that the designation of Anthropic as a supply-chain risk was illegal (59-page decision). (theguardian.com ; techcrunch.com) | 2026-08-31 — Pentagon launched versions of ChatGPT and Grok for 3 million civilian and military personnel (TechCrunch); GenAI.mil counts 1.7 million users among the 3 million workforce (Fortune). (techcrunch.com ; fortune.com)
Follow the trail
Walk this on the live map →
Part of the Psychohistory engine — 2,874 entities, 7,526 documented connections. Open data, built to be proven wrong.