행정에서 인공지능의 활용에 있어서 행정법적 문제과 개혁과제

Abstract

In order to achieve a fair, efficient, rapid and predictable exercise of state power, it is necessary to prepare for the emergence of a “government for the people, using artificial intelligence, for the people,” which has jumped from a “government for the people, by the people, for the people” by accepting advanced science and technology and reflecting the changed social image, so it is necessary to review the active use of artificial intelligence within the framework of the rule of law administration, check administrative law, and improve legislation. Analysis of the use of artificial intelligence in administrative law does not stop at deriving the current law's interpretation or legislative tasks for the use of artificial intelligence under the Administrative Procedure Act or limited to administrative disposition, but also checks the public law interpretation and response to the use of artificial intelligence in all “administrations” subject to administrative law. Criticism or problem consciousness of current artificial intelligence mainly deals with information opacity, process inexplicability, non-decipherability, information contamination, and intentional distortion of information in the deep learning stage and the use of generative artificial intelligence. The general legislative attitude of the EU and others has attempted to establish a legislative safety net to prevent the emergence of market economy monsters using artificial intelligence, but it should be taken into account that policy judgment and normative consideration are needed to prepare for the emergence of information monsters at the same level as public and private areas, and discussions on risks are focused on deep learning-based artificial intelligence. The EU-type behavioral regulation model regulates general behavior in the use of artificial intelligence regardless of the public and private areas, but private economic entities have no democratic legitimacy in the organizational composition and pursue the maximum of private interests by nature. It is overlooked that the administrative entity, which is the public authority, has democratic legitimacy in the composition, and that the possibility of stakeholder participation in administrative processes and specific administrative procedures is guaranteed, as well as checks by the principle of separation of powers. If the act regulation law in the use of artificial intelligence by private economic entities is brought into the public domain, it leads to overregulation in the public domain, but underregulation for private economic entities occurs. Data is the resource of the future and the oil of the future in the age of artificial intelligence. The data must be managed publicly and the public's free and stable use must be guaranteed if the public property of the data is recognized and pseudonymized information is released from the personal personal rights and ownership of the data. A department in charge of creating, collecting, and managing big data at the national level will be essential. Hangeul data, with the originality of Hangeul, needs to approach the protection of Hangeul resources and the development of Hangul-specific artificial intelligence in order to protect national interests. It is necessary to actively utilize artificial intelligence to increase the objectivity and rationality of evaluation in evaluating business feasibility or predicting future demand in various private capital attraction projects and public development projects, and it is necessary to improve the legal system that detects the misappropriation and overcharges of public funds, imposes punitive fines, and the use of artificial intelligence in the process of selecting subjects and verifying them through post-mortem review in the public benefit area.