How Is the AI Act Implemented and Enforced?

The AI Act contains a comprehensive enforcement regime, which essentially consists of two parts. First, high-risk AI systems are tested in a conformity assessment procedure before being placed on the market or put into service (see Sect. 6.1). Second, all AI systems and GPAI models are monitored by surveillance authorities, particularly after they have been made available on the market (see Sect. 6.2). To further facilitate and promote compliance with the AI Act, the AI Act also contains additional requirements to test AI systems under real-world conditions before they are put into service or placed on the market (see Sect. 6.3).