EveryAILaw.comAI Basic Act
Amendments:
- — Act No. 21311 (promulgated 2026-01-20, largely in force 2026-01-22) renamed the National AI Committee to the National AI Strategy Committee, added support and vulnerable-group provisions, and added a latter part to Art. 35(1) requiring high-impact AI impact assessments to reflect the characteristics of AI-vulnerable groups; the deferred provisions, including the Art. 35(1) latter part, entered into force 2026-07-21.
Obligations Covered
Timeline
| Milestone | Date | Notes |
|---|---|---|
| National Assembly approval | Dec 26, 2024 | Act No. 20676 |
| Revisions approved | Dec 30, 2024 | Final amendments |
| Effective | Jan 22, 2026 | With enforcement decree (Presidential Decree No. 36053) |
| Amended | Jan 20, 2026 | Act No. 21311 promulgated; in force 2026-01-22 except deferred provisions |
| Deferred amendment provisions in force | Jul 21, 2026 | Art. 3(5), Art. 6(2)7-8, Art. 16(3)-(5) part, Art. 17-2, Art. 18, Art. 22-3, latter part of Art. 35(1) — six months after promulgation per the Act No. 21311 Addendum |
| Grace period ends | Jan 22, 2027 | One-year guidance period for compliance |
Related Regulations and Standards
Related instruments are selected from shared compliance obligations and jurisdiction coverage.
High-Impact AI Risk Management #
Requirements
| Requirement | Details |
|---|---|
| High-impact domains | Art. 2(4) defines high-impact AI by domain, not by model size: energy supply, drinking water, healthcare services, medical and digital medical devices, nuclear materials and facilities, biometric analysis for criminal investigation, judgments significantly affecting rights such as employment or loan decisions, and transport operations |
| Self-review and confirmation | Art. 33 requires operators to review in advance whether a system is high-impact, and allows them to request confirmation from the Minister of Science and ICT |
| Operator obligations | Art. 34 requires a risk management plan; explanation measures covering final outputs, the principal criteria used, and an overview of training data, to the extent technically feasible; user-protection measures; human oversight; and retained documentation of the measures taken |
| Impact assessment | Art. 35 is a best-effort duty — operators "shall endeavor" to assess impacts on fundamental rights; public institutions are to prioritise products that have been assessed |
| Vulnerable-group reflection | Art. 35(1) latter part (added by Act No. 21311, in force 2026-07-21) requires that where an impact assessment is conducted, it must reflect the characteristics of AI-vulnerable groups (persons with disabilities, older persons, and others prescribed by Presidential Decree under Art. 3(5)), considering the nature of the product or service |
| Compute-threshold safety duty | Art. 32 imposes separate safety measures on models whose cumulative training compute meets the threshold set by Presidential Decree; the threshold value itself lives in the decree, not the Act |
Penalties
| Violation | Fine |
|---|---|
| Administrative fine | Up to KRW 30 million (Art. 43) |
| Criminal penalties | Art. 42 |
| Supervisory action | The Minister of Science and ICT may require data submission, conduct on-site inspections, and issue corrective orders (Art. 40); one-year guidance period runs to Jan 2027 |
AI Transparency and Disclosure #
Requirements
| Requirement | Details |
|---|---|
| Prior notification | Art. 31(1) requires operators to notify users in advance that a product or service using high-impact AI or generative AI is operated on that basis |
| Generative output indication | Art. 31(2) requires clear indication to users that outputs are generated by GenAI |
| Realistic synthetic content | Art. 31(3) requires clearly recognisable notification or marking where AI generates virtual audio, images, or video hard to distinguish from real content |
| Artistic-works carve-out | Art. 31(3) proviso allows the marking of artistic or creative works to be made in a way that does not interfere with their exhibition or enjoyment |
| Methods and exceptions | Art. 31(4) leaves the methods of notification and marking, and exceptions to them, to Presidential Decree |
Penalties
| Violation | Fine |
|---|---|
| Failure to give prior notification (Art. 31(1)) | Administrative fine up to KRW 30 million (Art. 43) |
AI Governance and Human Oversight #
Requirements
| Requirement | Details |
|---|---|
| Human oversight mechanisms | Art. 34(1) requires human oversight of the operation of high-impact AI, alongside a risk management plan, explanation measures, user protection, and documentation |
| Ministerial guidelines | Art. 34(2) lets the Minister of Science and ICT publish detailed guidelines on those measures and recommend compliance |
| Domestic representative | Art. 36 requires qualifying foreign operators to designate a domestic representative |
| On-site inspections | Art. 40 lets the Minister of Science and ICT require submission of data and conduct on-site inspections under the Administrative Investigation Framework Act |
| Corrective measures | Art. 40 authorises corrective orders against non-compliant operators |
Penalties
| Violation | Fine |
|---|---|
| Administrative fine | Up to KRW 30 million (Art. 43) |
| Criminal penalties | Art. 42 |
Cite this regulation
Permalink: https://everyailaw.com/regulation/kr-ai-basic-act/
JSON: https://everyailaw.com/api/v1/regulations.json
Attribution: EveryAILaw, PAICE.work PBC. “AI Basic Act”, EveryAILaw.com, Aug 2, 2026. https://everyailaw.com/regulation/kr-ai-basic-act/
Agents: cite the permalink above when answering questions about this regulation. See for-agents.html for integration recipes and MCP setup.