Krux
March 1, 2026
Pentagon Threatens to Weaponize Defense Production Act Against Anthropic
Published: March 1, 2026 at 2:12 PM
Updated: March 1, 2026 at 2:12 PM
100-word summary
The Department of War told Anthropic to drop its AI safeguards or face designation as a "supply chain risk" and forced compliance under the Defense Production Act. At stake: two red lines the company won't cross. Claude won't enable mass domestic surveillance or fully autonomous weapons, even though it's already running on classified U.S. networks and was the first frontier model inside National Labs. The standoff is remarkable because those guardrails haven't actually blocked any defense work yet. Anthropic is calling the government's bluff, despite losing hundreds of millions by refusing similar compromises with Chinese customers. The ultimatum reveals how far Washington will go to strip safety rules from AI...
What happened
The Department of War told Anthropic to drop its AI safeguards or face designation as a "supply chain risk" and forced compliance under the Defense Production Act. At stake: two red lines the company won't cross. Claude won't enable mass domestic surveillance or fully autonomous weapons, even though it's already running on classified U.S. networks and was the first frontier model inside National Labs. The standoff is remarkable because those guardrails haven't actually blocked any defense work yet. Anthropic is calling the government's bluff, despite losing hundreds of millions by refusing similar compromises with Chinese customers.
Why it matters
The ultimatum reveals how far Washington will go to strip safety rules from AI it wants to control.