Autonomous targeting without human control and untargeted profiling form the core prohibitions of this new oversight framework. Designed for Google Cloud contracts, the system balances operational security with strict auditing. It aims to close loopholes that typically allow government deals to bypass safety restrictions. Practitioners can now see a concrete model for enforceable AI constraints.