On 9 December 2023, the Parliament and the Council of the European Union adopted a provisional agreement on the proposal on harmonised rules on artificial intelligence (AI Act). The compromise agreement clarifies the definition of an AI system by aligning it with the OECD's proposed approach, seeking to differentiate AI from simpler software systems. The AI Act specifically exempts systems exclusively utilised for military or defence purposes and those employed solely for research, innovation, or non-professional personal use. The agreement includes the establishment of a European AI Board and AI Office within the Commission. The AI Office will be assigned the tasks of monitoring and ensuring compliance with the Act, conducting joint investigations, collaborating with national supervisory authorities, and offering guidance on the Act's obligations. A scientific panel of independent experts will contribute to developing methodologies for evaluating foundation models, providing recommendations on high-impact foundation models, and monitoring potential material safety risks associated with such models. Furthermore, the Act includes a framework for developing AI regulatory sandboxes. The sandboxes, designed for the controlled development, testing, and validation of innovative AI systems, will allow testing in real-world conditions. Following the provisional agreement, the next steps involve finalizing the document and submitting it to the European Union Parliament for adoption.
Original source