Reasons for Proposal

Artificial intelligence (AI) is used worldwide in various fields such as transportation, medical care, and other services, raising expectations for the possibility of industrial utilization as well as the improvement of convenience in daily life. 

It is expected that the development and use of AI will be applied to almost all industrial fields to create added value, but on the other hand, concerns about the dangers of AI are also growing. In particular, AI technology is based on machine learning that improves performance by learning large amounts of data, so it has uncertainty and opacity and a high possibility of generating errors due to noise data. 

Therefore, there is a need for legal regulation on the matter that AI technology can be deliberately abused beyond the level of human control in a specific field. Since the release of ChatGPT 4.0, it has become more urgent to establish “guidelines” through legislation on what is and is not permitted. Considering the situation where the European Parliament is scheduled (around 2025) to enact the Artificial Intelligence Act for the first time in the world, Korea also needs to accurately identify the side effects and risks of AI and form AI regulatory policies. The goal of AI-related regulations should be to move toward a society where sustainable growth is possible beyond simple regulations on the risks of new technologies, avoiding suppression or restrictions on individual freedom. 

Therefore, to create a trustworthy AI use environment in the future, this Bill aims to create an institutional foundation for safe and reliable AI technology and policies, such as defining the basic principles for the development and use of AI, the responsibilities of AI business operators, and the rights of users and separating and providing policies related to AI, to protect users by classifying the types of AI into prohibited AI, high-risk AI, and low-risk AI.

Major Provisions

A. The purpose of this Act is to protect the rights, interests, and dignity of citizens and to improve their quality of life by creating a foundation for the development/use of safe and reliable AI and prescribing matters necessary for the establishment/promotion of AI-related policies (Article 1).

B. Define AI, prohibited AI, high-risk AI, low-risk AI, and AI business operators (Article 2).

C. Establish the principle of preferential permission/ex post regulation for the development and use of AI other than prohibited AI (Article 5).

D. Prohibit the development of prohibited AI in principle (Article 6).

E. Permit the development and use of low-risk AI in principle, but in cases of interacting by detecting the user's biometric data, or creating a photo, audio, video, etc., as real, make that fact disclosed (Article 7).

F. Stipulate the role of the government, the responsibilities of business operators, and the rights of users to protect themselves from high-risk AI (Articles 8 through 11).

G. For safe and reasonable development and use of AI, require the Minister of Science and ICT to establish and implement a master plan for AI every three years through deliberation and resolution by the Artificial Intelligence Committee, and ensure that the master plan includes matters related to the basic direction and strategy of AI-related policies, and the creation of a foundation of trust (Article 12).

H. Require the establishment of the Artificial Intelligence Committee under the Prime Minister in order to deliberate and decide on matters related to the realization of an AI society and securing trust in the AI industry, and let the Artificial Intelligence Committee deliberate/resolve matters related to the establishment of the master plan and the inspection/analysis of its progress (Articles 13 and 14).

I. Require the Minister of Science and ICT to come up with policies to minimize the potential risks of AI, etc. to citizens’ lives and create a foundation of trust for the safe use of AI (Article 19).

J. Establish an identification system for prohibited AI and high-risk AI (Articles 21 through 23).

K. Require high-risk AI business operators to take measures to ensure the reliability and safety of AI (Article 24).