The applications of AI systems must be controlled and managed for the benefit of all, because of their potential influence on the contextual environment. This Chapter presents an AI risk management framework which, while addressing compliance with the European AI Act, provides consistent and rigorous assessment metrics to measure and monitor the risk that AI may cause on individuals, organisations, or the planet at large, thereby balancing opportunities with risks. The Chapter argues that the principles of Security, Accuracy, Fairness, and Explainability (SAFE) for AI systems are the basis for an effective risk management framework and provides practical guidelines for its implementation.