Policy-orientated AI is a term used in the fields of artificial intelligence, industry and Factory 4.0 as well as cybercrime and cybersecurity. It describes artificial intelligence (AI) that works according to clearly defined rules and specifications. This means that the AI's decisions are based on defined guidelines - for example, legal requirements, ethical guidelines or specific company rules.
Instead of just learning from large amounts of data, a guideline-orientated AI combines existing knowledge with established rules. This ensures that AI systems become more predictable and secure - especially in sensitive areas such as production, intelligent manufacturing or the monitoring of IT systems.
One example: In a factory that works with AI-controlled robots, a guideline-orientated AI ensures that robots maintain certain safety distances from people and do not perform any dangerous movements. Even if the AI is otherwise very flexible, it never exceeds these defined limits.
Policy-oriented AI helps companies to utilise innovations without losing sight of important rules.
Deutsch 
English (UK)