US appeals court reportedly upholds Pentagon’s Anthropic supply-chain risk designation
A user says Claude’s built-in refusals of certain military tasks drove the decision—not a compromised model—and argues it raises questions about India’s reliance on foreign AI.
TLDR
In a September 27, 2026 post, a user says a US appeals court upheld the Pentagon’s decision to treat Anthropic as a supply-chain risk. The author says the court accepted that Claude’s built-in limits could make it unreliable for defense work. The post frames this as a question of “model sovereignty” for India: who controls what AI refuses to do in sensitive systems, and whether to rent the best model it cannot fully control or accept a slower one it owns.
Combined views
—
1 Source, first seen 3h ago
