On 22 January 2026, the Ministry of Science and ICT released the high-impact artificial intelligence (AI) operator obligations guidelines to support the implementation of Article 34 of the Artificial Intelligence Basic Act. The guidelines elaborate on the statutory responsibilities of operators that provide or use high-impact artificial intelligence products or services. They specify obligations relating to the establishment and operation of risk management measures, explanations of artificial intelligence outputs and decision criteria within technically feasible limits, disclosure of information on training data used in development and operation, user protection measures, human oversight mechanisms, and the preparation and retention of documentation demonstrating safety and reliability measures. The Guidelines explain how these obligations apply across the full lifecycle of high-impact artificial intelligence and clarify their role as interpretative guidance supporting compliance with existing legal obligations rather than creating additional binding requirements.
Original source