Luckey backs Pentagon's Anthropic crackdown

This is the sharpest test yet of whether AI companies can set ethical limits on government use of their models — and the early answer from the Pentagon is: no.

• –The DOD's supply-chain risk label has historically been reserved for foreign adversaries; applying it to a U.S. company over usage-policy disagreement sets a legally and politically explosive precedent
• –Anthropic drew a specific red line — no fully autonomous weapons, no domestic mass surveillance — while the Pentagon demanded unfettered access across all lawful purposes; neither side blinked
• –Luckey's position is nakedly self-serving: Anduril just landed a $20B Army contract and competes directly in the defense AI space that Anthropic is now locked out of
• –Anthropic's lawsuit will likely become a landmark on government AI procurement policy and whether agencies can blacklist vendors who impose safety constraints
• –For developers building on Claude for defense or federal contractors, this creates immediate supply-chain uncertainty — certifying non-use of Anthropic models may now be a contract requirement