What is Prohibited-practice screening?

In practice, this is a front-end gate in the AI intake or approval workflow. Compliance teams review the intended purpose, user group, data sources, and operational behavior to determine whether the system would fall into a banned category such as certain manipulative, exploitative, or unlawful biometric uses, depending on the jurisdiction and policy set. The output should be a clear allow, reject, or escalate decision with supporting evidence.

For compliance, screening reduces the risk of investing in a system that cannot be lawfully deployed and helps create a documented defense for procurement and launch decisions. The concept is most directly tied to the EU AI Act prohibition regime, but similar screening logic is used in internal AI governance programs, AI management systems, and risk assessments for sectors with stricter conduct expectations.