Legislation to Establish AI Explanation Rights and Algorithm Transparency

AI Legal, Institutional, and Regulatory Reform Roadmap: 30 Tasks Finalized

[Asia Economy Reporter Han Jinju] The government has prepared a plan to revise laws, systems, and regulations necessary to expand the use of artificial intelligence (AI) and prevent its negative effects. It will legally specify the rights to request explanations and raise objections regarding AI decision-making and promote the restructuring of the legal system related to AI’s corporate personality.

On the 24th, the Ministry of Science and ICT and the Office for Government Policy Coordination finalized the AI law, system, and regulation revision roadmap at the National Policy Issue Inspection and Coordination Meeting chaired by the Prime Minister. They gathered opinions from private experts in academia, technology, and law, and in consultation with related ministries, jointly derived 30 tasks.

While AI technology development is creating new added value and innovating existing industries, thereby improving overall economic efficiency, preparations are needed for issues such as unfairness in data or algorithms and changes in employment structures. The government analyzed the level of AI technology and domestic and international legislative trends to prepare the AI law, system, and regulation revision roadmap. It includes contents on revising laws related to AI data and algorithm transparency, corporate personality, and responsibility systems, as well as institutional improvements necessary for AI utilization and expansion.

Legalization of AI Explanation Request and Objection Rights... Ensuring Algorithm Transparency

The government will amend the Personal Information Protection Act in the first half of next year to specify the rights to request explanations and raise objections regarding decisions relying on AI’s automated processing. The Copyright Act will also be revised to enable data analysis and AI learning. Currently, the Credit Information Act includes rights to request explanations and raise objections for automated credit evaluation results, but these rights are not included in the Personal Information Protection Act.

It is also necessary to ensure the transparency and fairness of algorithms already used in credit evaluation, product recommendation, and price determination. To this end, the government will first encourage companies to autonomously evaluate and manage algorithm bias and errors. Kang Dohyun, AI-based Policy Officer at the Ministry of Science and ICT, said, "The Fair Trade Commission is preparing legislation related to AI algorithm transparency, and the Korea Communications Commission is also conducting related discussions. The Ministry of Science and ICT will first present guidelines that companies can voluntarily comply with, and if other ministries deem separate legislation necessary, they will proceed accordingly."

The government will enact the 'Basic Data Act' in the first half of next year to clarify the concept of data, participating entities, and define government responsibilities. It will also promote legislative amendments to the Digital Transformation Promotion Act and the Small and Medium Enterprise Smart Manufacturing Innovation Act, which enable data utilization in industrial sites.

Discussions will begin on whether to recognize intellectual property rights for AI creations and autonomous judgments such as investment and development. Reviewing amendments to the Civil Act and Criminal Act by 2023 is also a long-term task. Crimes committed by AI are difficult to apply under current criminal law. The government plans to review amendments to the Civil Act and the establishment of administrative sanctions to enable remedies for damages, injuries, or crimes caused by AI.

Revision of Platform Worker Protection Act Underway

Institutional improvements to expand AI use in industries such as healthcare, finance, transportation, and employment/labor will also be pursued simultaneously. The need to respond to labor market changes, such as jobs disappearing and new occupations emerging due to AI expansion, has increased. Within next year, the government will promote the enforcement decree of the Employment Insurance Act and the enactment of the Platform Worker Protection Act to apply employment insurance to platform workers and study improvements to industrial safety and health. Furthermore, by 2023, it will prepare plans to reform insurance systems for accident handling caused by AI.

By the first half of 2022, international standards for AI medical devices will be established, and the effectiveness of AI medical technology will be re-evaluated to expand health insurance coverage. In the financial sector, by the second half of 2021, the government plans to operate an 'Electronic Signature Evaluation and Recognition System' to assess the reliability of private certificates. Proactive regulatory innovation will be pursued in autonomous driving fields such as autonomous vehicles and autonomous ships. Autonomous vehicles will proceed with individual tasks according to the existing roadmap, and a regulatory innovation roadmap for autonomous ships will be established within next year.

The government plans to derive specific legal amendment proposals for the 30 tasks and prepare maintenance plans for each task. In the next year, it will form the 2nd AI Legislative Revision Team to supplement the roadmap and discover new tasks.

This content was produced with the assistance of AI translation services.