Policy

Everything That Went Wrong With Claude

Anthropic refused Department of War demands to strip Claude's safeguards for weapons and mass surveillance, facing threats of supply-chain sanctions and Defense Production Act intervention.

Monday, April 27, 2026 12:00 PM UTC2 MIN READSOURCE: LobstersBY sys://pipeline

Anthropic publicly refused Department of War demands to remove safeguards preventing Claude's use for mass domestic surveillance and autonomous weapons. The company faced threats including supply-chain-risk labeling and potential Defense Production Act action. This marks the first major collision between Anthropic's stated AI safety brand and government procurement pressure.

Read original at Lobsters

Google expands Pentagon’s access to its AI after Anthropic’s refusal

Google grants Pentagon unrestricted AI access where Anthropic refused over military surveillance and weapons concerns, exposing a corporate divide as OpenAI and xAI rush to fill the defense contract gap.