The dispute between the AI company and the Department of War raises key questions about who should control AI, and how.

After signing a deal with the Department of War, OpenAI tries to build trust by publishing the contract details. So far, it's not working.

Only Congress can put meaningful limits on government abuse of AI.