South Korea - AI Development Framework

Bill on Artificial Intelligence (Representative: Lee Yong-bin, proposed July 19, 2021)

인공지능법안 (대표: 이영빈, 2021년 7월 19일 제안)

South Korea

RAI-KR-NA-AIRLYXX-2021
Withdrawn(Withdrawn)
BillGovernance and OversightAccountability and Documentation
Export PDF

Proposed on 19 July 2021 by Representative Lee Yong-bin (이용빈), the 'Bill on Artificial Intelligence' aimed to establish a framework for national AI policy, industry promotion, ethical safeguards, and coordination across central and local government. The bill sought to set definitions, institutional roles, duties for public and private actors, and transparency and data governance principles, but it did not become law and was not enacted (expired at the end of the 21st National Assembly term).

Overview

The Bill on Artificial Intelligence (의안명: Bill on Artificial Intelligence) was introduced to the National Assembly of the Republic of Korea as a member-drafted bill by Representative Lee Yong-bin (이용빈) on 2021-07-19. The proposal was presented as a framework statute intended to promote the AI industry, strengthen national competitiveness in artificial intelligence, and set out measures and roles for central and local governments to support AI development and deployment. The member bill was one of several AI-related legislative initiatives during the 21st National Assembly session. Contemporary reporting and parliamentary summaries of the member bill emphasize that the proposal primarily sought administrative and promotional mechanisms rather than prescriptive, system-level technical regulation. (See contemporaneous reporting: https://mobile.newsis.com/view/NISX20210719_0001518028?utm_source=openai and parliamentary summaries: https://kind.krx.co.kr/external/2025/07/28/000618/20250728001523/10601.htm?utm_source=openai).

Definitions

Public summaries of the draft indicate the bill included a statutory definition of "artificial intelligence" for the purposes of the instrument, intended to provide definitional clarity for implementing the proposed promotional and coordination measures. Available legislative summaries for the member bill do not record extensive technical subclassifications (for example, categories such as "high‑risk AI") with the prominence seen in later AI bills; instead, the focus in the member-drafted text (as summarized publicly) was on enabling policy measures rather than detailed technical categorization. For further detail in contemporary reporting and parliamentary tracking see: https://kind.krx.co.kr/external/2025/07/28/000618/20250728001523/10601.htm?utm_source=openai.

Governance and Institutional Framework

The bill proposed a governance approach that emphasized central coordination and intergovernmental cooperation to deliver AI promotion and policy measures. Summaries describe central government roles for fostering AI industry growth and supporting an AI ecosystem, together with coordinating responsibilities among national ministries and local governments. The draft appeared to rely on existing governmental bodies rather than creating a large standalone regulatory agency or a novel enforcement authority in the text of the member bill; the emphasis was state-led facilitation and institutional coordination to advance R&D, infrastructure and workforce initiatives. (Contemporaneous reporting and legislative trackers: https://kind.krx.co.kr/external/2025/07/28/000618/20250728001523/10601.htm?utm_source=openai; https://mobile.newsis.com/view/NISX20210719_0001518028?utm_source=openai.)

Key Focus Areas

  • Government responsibilities for AI industrial promotion and ecosystem support — establishing duties and roles for central and local public authorities to encourage AI growth and competitiveness.
  • Inter-ministerial and local coordination — setting principles and mechanisms for coordination among ministries and local governments to deliver coherent AI promotion policies.
  • Policy measures for R&D, infrastructure, and workforce development — enabling public programs and support measures to strengthen national AI capabilities.
  • Standardization and institutional support — promoting standardization activities and institutional measures to facilitate AI development, rather than prescribing technical obligations on individual systems.
  • Promotion and facilitation emphasis — the draft prioritized measures to promote the AI industry and support ecosystem actors, with less emphasis on detailed prohibitions, system-level risk controls or private-sector liability rules.

Implementation Framework

The draft law was framed to apply across public and private sectors insofar as necessary to implement state and local government AI promotion policies and measures. Its scope, according to public summaries, emphasized sectoral support activities (research funding, infrastructure development, skills and workforce programs) and administrative coordination rather than a comprehensive regime of duties, prohibitions or technical compliance obligations for individual AI systems or private providers. Because the draft was promotional and coordinative in character, the compliance requirements recorded in public summaries centered on administrative implementation tasks for public bodies — planning, reporting and intergovernmental cooperation — and the text as summarized did not set out prescriptive obligations such as mandated notices, automated-decision safeguards or extensive private-sector liability provisions. If the bill had proceeded through the legislative process and been enacted, more detailed obligations and operational rules likely would have been specified in implementing regulations or subordinate instruments. (See legislative summaries and reporting: https://kind.krx.co.kr/external/2025/07/28/000618/20250728001523/10601.htm?utm_source=openai; https://mobile.newsis.com/view/NISX20210719_0001518028?utm_source=openai.)

Monitoring and Evaluation

Publicly available summaries of the member bill do not set out an extensive monitoring, evaluation, or oversight architecture for independent review of AI systems or centralized audit mechanisms. The draft’s monitoring and evaluation emphasis, as described in summaries, would likely have focused on administrative monitoring of the implementation of promotional programs and intergovernmental coordination — for example, planning and reporting duties of public bodies — rather than risk‑based monitoring of private AI systems. Detailed arrangements for monitoring program effectiveness and for evaluating policy outcomes were not recorded in the publicly available summaries and would likely have been addressed through implementing measures and administrative practice if the legislation had advanced. (Contemporaneous summary reporting: https://kind.krx.co.kr/external/2025/07/28/000618/20250728001523/10601.htm?utm_source=openai.)

Penalties, Liability, and Appeals

Public summaries and contemporaneous reporting do not identify substantive new enforcement regimes, novel penalties, or expanded private‑sector liability measures contained in the member-drafted bill. The legislative emphasis was on institutional measures to promote AI (coordination, program support, infrastructure), and enforcement or penalty mechanisms — if any — were not prominent in the public descriptions of this draft. As a promotion‑focused statute, the draft did not, in the available summaries, articulate detailed appeals or administrative sanction procedures directed at AI system operators in the way that later, risk-focused proposals sometimes do. Any specific penalty, liability or appeal structures would likely have been specified in implementing rules or distinct regulatory instruments had the bill progressed. (Sources: https://kind.krx.co.kr/external/2025/07/28/000618/20250728001523/10601.htm?utm_source=openai.)

Relationship to Other Instruments

The Lee Yong‑bin member bill was one of multiple AI‑related legislative proposals introduced during the 21st National Assembly. It should be understood in context with other member and government bills addressing AI promotion, ethics, standards and risk‑based regulation. Subsequent legislative activity in later sessions continued to develop AI law themes — including promotion, trust and risk management — and later packages and government proposals consolidated or superseded many member initiatives. The 2021 proposal did not become the principal instrument that defined Korea’s later comprehensive AI framework and was recorded in legislative compilations as having lapsed at the end of the parliamentary term. (See parliamentary tracking and summary references: https://kind.krx.co.kr/external/2025/07/28/000618/20250728001523/10601.htm?utm_source=openai; https://kind.krx.co.kr/external/2025/06/05/000458/20250605001236/10001.htm?utm_source=openai.)

International Alignment

The public summaries of this member bill do not address specific international alignment or cross‑border regulatory coordination measures (for example, harmonization with EU, OECD or other international AI frameworks) within the text as summarized. The primary focus in the available descriptions was on national promotion, interministerial coordination and support measures. Stakeholders interested in international comparability would need to consider later legislative developments and implementing regulations that explicitly address international standards, data flows or cross‑border regulatory cooperation. (No specific international instruments were cited in the public summaries of this member-drafted bill.)

Implementation Timeline

DateEvent
2021-07-19Bill introduced (representative filing) by Representative Lee Yong-bin and registered/entered the parliamentary system.
2021-07-20Referred to committee (summary entry in parliamentary trackers).
2024-05-29Legislative term expiration — bill lapsed/expired at the close of the 21st National Assembly term (임기만료 폐기); the proposal did not become law.

Compliance Checklist

RequirementDescription
Public authority planning and reportingThe draft centered compliance tasks on administrative planning, reporting and intergovernmental cooperation obligations for national and local public bodies to implement AI promotion measures.
Intergovernmental coordinationEstablish or follow principles for coordination among ministries and local governments when delivering AI promotion programs and infrastructure projects.
Programmatic support implementationImplement R&D, infrastructure, standardization and workforce development programs consistent with the promotional objectives set out in the draft.
No detailed private‑sector technical compliance in draftPublic summaries indicate the member bill did not impose detailed prescriptive obligations on private AI system operators (e.g., automated-decision notices, risk-classification duties); such obligations were left to other legal regimes or potential future implementing rules.

Sources and References

SourceURL
National Assembly — Bill search (의안정보시스템)https://likms.assembly.go.kr/bill/billList.do
Plain English

The South Korean "Bill on Artificial Intelligence," introduced in July 2021, aimed to establish a national framework for promoting the artificial intelligence industry and coordinating government efforts, but it ultimately did not become law. This proposal primarily outlined duties for central and local government bodies to foster AI development and strengthen national competitiveness, rather than imposing direct technical regulations or prohibitions on private companies.

The bill’s scope centered on public sector responsibilities. It sought to define AI for policy purposes and create institutional roles for government agencies. Key proposed obligations for public authorities included: - Promoting AI industry growth and supporting a robust AI ecosystem. - Ensuring inter-ministerial and local government coordination for coherent AI policies. - Enabling public programs for research and development, infrastructure, and workforce training. - Advancing standardization activities to facilitate AI development.

Unlike some later AI proposals, this bill did not introduce detailed prescriptive obligations for private AI system operators, such as specific risk classifications, mandated notices, or extensive liability rules. Its emphasis was on state-led facilitation and institutional coordination.

This bill never took effect. It was introduced on July 19, 2021, and referred to committee, but it lapsed and was withdrawn at the expiration of the 21st National Assembly term on May 29, 2024. Consequently, it did not introduce any new enforcement mechanisms, penalties, or private-sector liability measures. A practical surprise for readers is that despite its title, this bill was a promotional framework for government action, not a regulatory regime for private AI systems, and its failure to pass means it carries no current legal weight.

Plain-English rewrite by Regulations.ai — not legal advice. Verify against the official text.

What you must do — compliance checklist

0 / 6 marked complete

Plain-English obligations under South Korea - AI Development Framework. Not legal advice — verify against the official text before relying on it.

  1. #1Important

    Applies to: National and local public bodies

    The draft centered compliance tasks on administrative planning, reporting and intergovernmental cooperation obligations for national and local public bodies to implement AI promotion measures.
  2. #2Important

    Applies to: National ministries and local governments

    Establish or follow principles for coordination among ministries and local governments when delivering AI promotion programs and infrastructure projects.
  3. #3Important

    Applies to: Public authorities

    Implement R&D, infrastructure, standardization and workforce development programs consistent with the promotional objectives set out in the draft.
  4. #4Important

    Applies to: Central government bodies

    central government roles for fostering AI industry growth and supporting an AI ecosystem
  5. #5Important

    Applies to: National ministries and local governments

    coordinating responsibilities among national ministries and local governments
  6. #6Important

    Applies to: Public authorities

    promoting standardization activities and institutional measures to facilitate AI development

© Regulations.AI — created on 13-Jun-2026