AI policy
Who Gets to Tell the Robot No?
Anthropic built the most trusted AI in classified government networks — and then refused to let the Pentagon use it however it wanted. The fallout is reshaping the future of AI, warfare, and American democracy.