AI Basic Act: requirements, deadlines and compliance actions
Framework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust (AI Basic Act, Act No. 20676)
What this label means
- Verified
- — a named reviewer opened the official source and confirmed the record; the date and the reviewer are shown.
- Pending review
- — linked to the official source and in the queue for a reviewer to confirm.
- Source-linked
- — linked to the official source, not yet confirmed by a reviewer.
On this page 9 sections
In brief
AI Basic Act is a binding act of South Korea, issued by the National Assembly of the Republic of Korea; competent ministry Ministry of Science and ICT. It has been in force since 22 January 2026. Scope is as set out in the official text linked below. 9 obligations are recorded against it, each with its source reference.
- Jurisdiction
- South Korea
- Instrument type
- Act / statute
- Status
- In force
- Legal force
- Binding
- Issuing body
- National Assembly of the Republic of Korea; competent ministry Ministry of Science and ICT
- Adopted
- 21 January 2025
- In force
- 22 January 2026
- Applies from (first)
- 22 January 2026
- Obligations recorded
- 9
- Official source
- Framework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust (AI Basic Act, Act No. 20676)
- Verification
- Verified against the official source 26 Sep 2026
What is AI Basic Act?
Korea's AI Basic Act is a comprehensive framework statute combining industrial promotion with trust and safety rules. It gives a legal basis for a national AI master plan every three years and the National AI Committee, supports R&D, data, standardisation and SMEs, and imposes duties on operators of high-impact AI (risk management, explanation, human oversight) and transparency duties to notify users of generative AI output. Foreign operators above a threshold must designate a domestic representative. It was promulgated in January 2025 and took effect on 22 January 2026 after a one-year grace period.
Status note: Official page is in Korean; law number, promulgation date (2025-01-21) and enforcement date (2026-01-22) were confirmed from it. The 26 December 2024 plenary passage date comes from search results rather than the fetched page. English title is the commonly used translation; an official English text was not fetched.
Who does it apply to?
Scope is as set out in the official text linked below. Issued by National Assembly of the Republic of Korea; competent ministry Ministry of Science and ICT. Oversight: Ministry of Science and ICT as competent authority; presidential National AI Committee; AI Safety Institute. Stated purpose: Promote the AI industry while establishing trust obligations for high-impact and generative AI.
When do the requirements apply?
2024-12-26: Passed by the National Assembly plenary; 2025-01-21: Promulgated as Act No. 20676; 2026-01-22: Act enters into force
The in-force date is the one the key dates above record from the source; application dates are to be confirmed by a reviewer.
What must organisations do?
Each duty is one line: open it for the summary, practical action and evidence, or go to its own page.
Legal requirement AI business operators must notify users in advance that a product or service runs on high-impact or generative AI Provider / developer, Deployer / user organisation · from
An AI business operator that provides a product or service operated by high-impact AI or by generative AI must notify users in advance that the product or service is based on such AI. The notice is owed before use and is the entry point of the Act's transparency chapter; failure to give it can attract an administrative fine after a corrective order.
Practical action: Add a plain-language AI notice at onboarding or first use for any product that relies on high-impact or generative AI.
Evidence examples: User notice copy and placement record
Framework mapping (original, editorial): NIST AI RMF 1.0 GOVERN 5.1, MANAGE 4.1; ISO/IEC 42001:2023 Annex A.8.5
Open the obligation page →Article 31(1) Verified (a factual check against the official source, not a legal review)
Legal requirement AI business operators must label generative AI output and clearly flag realistic synthetic media Provider / developer, Deployer / user organisation · from
An AI business operator that provides generative AI or a product or service using it must indicate that the output was generated by generative AI. Where the operator provides content such as virtual sound, images or video that is hard to distinguish from reality, it must clearly notify users that the content is AI-generated, in a manner that does not impair a work's artistic or creative expression where that applies. Detailed methods are left to Presidential Decree and ministry guidance.
Practical action: Apply visible labels and, where feasible, embedded provenance marks to generated audio, image and video, and track the decree's technical specifications.
Evidence examples: Output labelling design and samples; Synthetic media labelling standard
Framework mapping (original, editorial): NIST AI RMF 1.0 MEASURE 2.7, MANAGE 4.1; ISO/IEC 42001:2023 Annex A.8.2, A.8.5
Open the obligation page →Article 31(2) and 31(3) Verified (a factual check against the official source, not a legal review)
Legal requirement Operators of AI above the compute threshold must run lifecycle risk management and report safety results Provider / developer, General-purpose AI model provider · from
An AI business operator whose AI system's cumulative training computation exceeds the threshold set by Presidential Decree must identify, assess and mitigate risks across the system's life cycle, build a risk management system to monitor and respond to safety incidents caused by the AI, and submit the results of these measures to the Minister of Science and ICT. The threshold and reporting form are set in the Decree.
Practical action: Track training compute against the decree threshold and prepare a safety report covering risk identification, mitigation and incident monitoring for submission to the ministry.
Evidence examples: Lifecycle risk assessment and mitigation record; Safety measures report submitted to the ministry
Framework mapping (original, editorial): NIST AI RMF 1.0 MAP 5.1, MEASURE 2.6, MANAGE 1.3, MANAGE 4.3; ISO/IEC 42001:2023 Clause 6.1.2, 6.1.3, 9.1
Open the obligation page →Article 32 Verified (a factual check against the official source, not a legal review)
Legal requirement Operators of high-impact AI must establish and operate a risk management plan Provider / developer, Deployer / user organisation +1 · from
An AI business operator that provides high-impact AI, or a product or service using it, must establish and operate a plan for managing the risks of that AI. High-impact AI is AI used in areas such as energy, drinking water, health care and medical devices, nuclear safety, biometric analysis for criminal investigation, decisions on recruitment or loans, transport, public services and education, where it may materially affect life, safety or fundamental rights.
Practical action: Classify systems against the high-impact areas and write a risk management plan for each one before launch.
Evidence examples: High-impact AI risk management plan; High-impact classification record
Framework mapping (original, editorial): ISO/IEC 42001:2023 Clause 6.1.2, 6.1.3, 8.1; NIST AI RMF 1.0 MAP 1.5, MANAGE 1.3
Open the obligation page →Article 34(1) Verified (a factual check against the official source, not a legal review)
Legal requirement Operators of high-impact AI must be able to explain outputs and the main criteria behind them Provider / developer, Deployer / user organisation +1 · from
Operators of high-impact AI must put in place measures to explain the AI's final results, the main criteria used to reach them, and an overview of the training data, to the extent that this is technically feasible. The duty targets explainability of the system's decisions to users and affected people rather than full disclosure of the model.
Practical action: Document the decision criteria and training-data overview for each high-impact system and prepare a user-facing explanation format.
Evidence examples: Explanation method and decision-criteria documentation
Framework mapping (original, editorial): NIST AI RMF 1.0 MEASURE 2.9, GOVERN 5.1; ISO/IEC 42001:2023 Annex A.8.2, A.7.3
Open the obligation page →Article 34(1) Verified (a factual check against the official source, not a legal review)
Legal requirement Operators of high-impact AI must ensure human management and supervision Provider / developer, Deployer / user organisation +1 · from
Operators of high-impact AI must ensure that the AI is subject to human management and supervision, so that people remain able to monitor its operation and intervene in or override its outputs where those outputs may affect life, physical safety or fundamental rights.
Practical action: Assign accountable supervisors for each high-impact system and design the interface so they can pause, correct or overrule outputs.
Evidence examples: Human supervision design and assigned supervisors
Framework mapping (original, editorial): ISO/IEC 42001:2023 Annex A.9.2; NIST AI RMF 1.0 GOVERN 3.2, MANAGE 2.4
Open the obligation page →Article 34(1) Verified (a factual check against the official source, not a legal review)
Legal requirement Operators of high-impact AI must prepare user-protection measures and keep records of their safety and trust measures Provider / developer, Deployer / user organisation +1 · from
Operators of high-impact AI must prepare and implement a plan to protect users, and must prepare and keep documents describing the measures they take to secure the AI's safety and reliability, so that they can be shown to the Ministry of Science and ICT on request. Further items may be added by Presidential Decree, and the ministry may publish guidelines on how to meet these duties.
Practical action: Write a user-protection plan for each high-impact system and keep a dated file of the safety and trust measures applied to it.
Evidence examples: User protection plan; Safety and reliability measures file
Framework mapping (original, editorial): ISO/IEC 42001:2023 Clause 7.5; Annex A.6.2.7, A.9.3; NIST AI RMF 1.0 GOVERN 1.4, MANAGE 4.1
Open the obligation page →Article 34(1) Verified (a factual check against the official source, not a legal review)
Voluntary Operators of high-impact AI should assess its impact on fundamental rights before use Provider / developer, Deployer / user organisation +1 · from
An AI business operator that provides high-impact AI, or a product or service using it, is to make efforts to assess in advance the impact the AI may have on people's fundamental rights. The provision is drafted as an endeavour duty rather than a hard requirement, but public bodies procuring high-impact AI are directed to give preference to products that have undergone such an assessment, and the ministry may set assessment methods.
Practical action: Run a fundamental-rights impact assessment for each high-impact system, particularly where public-sector procurement is in view.
Evidence examples: Fundamental-rights impact assessment for high-impact AI
Framework mapping (original, editorial): ISO/IEC 42001:2023 Clause 6.1.4; Annex A.5.2, A.5.4; NIST AI RMF 1.0 MAP 5.1, MAP 5.2
Open the obligation page →Article 35 Verified (a factual check against the official source, not a legal review)
Legal requirement Foreign AI business operators above the threshold must designate a domestic representative in Korea Provider / developer, General-purpose AI model provider +1 · from
An AI business operator without an address or place of business in Korea, whose user numbers or revenue meet the thresholds set by Presidential Decree, must designate in writing a domestic representative with an address in Korea. The representative handles the operator's duties under the Act, including submitting the Article 32 safety results, supporting high-impact AI confirmation and cooperating with fact-finding investigations, and the operator is treated as responsible for the representative's acts.
Practical action: Check user and revenue figures against the decree thresholds and appoint a Korean representative by written designation before the duty bites.
Evidence examples: Written designation of domestic representative
Framework mapping (original, editorial): ISO/IEC 42001:2023 Clause 5.3; Annex A.10.2; NIST AI RMF 1.0 GOVERN 2.1
Open the obligation page →Article 36 Verified (a factual check against the official source, not a legal review)
Official sources
-
Framework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust (AI Basic Act, Act No. 20676)
Korea Ministry of Government Legislation, National Law Information Center · 21 Jan 2025 · Tier 1 source
Cite this record
AIPolicyTracker (2026). “AI Basic Act” [AIPT-POL-0175, record version 2, verified 26 September 2026 by Bhaskar Bhatt]. https://aipolicytracker.org/id/AIPT-POL-0175 (accessed 27 October 2026). Data licensed CC BY 4.0.
Permanent id AIPT-POL-0175: it never changes or gets reused, and its address keeps working if this page moves.
Cite the official text alongside it: Framework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust (AI Basic Act, Act No. 20676), Korea Ministry of Government Legislation, National Law Information Center, https://www.law.go.kr/lsInfoP.do?lsiSeq=268543.
BibTeX
@misc{aipolicytracker2026_aibasicact,
author = {{AIPolicyTracker}},
title = {{AI Basic Act}},
year = {2026},
howpublished = {\url{https://aipolicytracker.org/policies/south-korea-framework-act-on-the-development-of-artificial-intelligence-and-establishment-of-a-foundation-for}},
url = {https://aipolicytracker.org/policies/south-korea-framework-act-on-the-development-of-artificial-intelligence-and-establishment-of-a-foundation-for},
urldate = {2026-10-27},
note = {Data licensed CC BY 4.0. Accessed 27 October 2026}
}
Using this in a product? Commercial use is allowed under CC BY 4.0, with attribution. Ask about an SLA-backed feed or a white-label licence.
Add a status badge to your site or README
It shows the status and review state as recorded here and updates when the record does.
Markdown
[](https://aipolicytracker.org/id/AIPT-POL-0175)HTML
<a href="https://aipolicytracker.org/id/AIPT-POL-0175"><img src="https://aipolicytracker.org/badges/south-korea-framework-act-on-the-development-of-artificial-intelligence-and-establishment-of-a-foundation-for.svg" alt="AIPolicyTracker: Framework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust (AI Basic Act, Act No. 20676), In force · verified 26 Sep 2026" height="20"></a>Change history
All updates for South Korea · RSS · Updates hub
Record history
-
- In force: — → 2026-01-22
Record version 2: Short title changed to "AI Basic Act", the name the official title itself gives, so pages and search results name the law the way people refer to it. No facts changed. Every earlier wording is in the GitHub repository.
Frequently asked questions
Is AI Basic Act in force?
Yes. AI Basic Act is in force since 22 January 2026.
Is AI Basic Act legally binding?
Yes. AI Basic Act is a binding act / statute; the obligations it creates are legal requirements for the actors it covers.
Who does AI Basic Act apply to?
Scope is as set out in the official text linked below. Issued by National Assembly of the Republic of Korea; competent ministry Ministry of Science and ICT. Oversight: Ministry of Science and ICT as competent authority; presidential National AI Committee; AI Safety Institute. Stated purpose: Promote the AI industry while establishing trust obligations for high-impact and generative AI.
When does AI Basic Act apply?
2024-12-26: Passed by the National Assembly plenary; 2025-01-21: Promulgated as Act No. 20676; 2026-01-22: Act enters into force
Where is the official text of AI Basic Act?
The official text is published by Korea Ministry of Government Legislation, National Law Information Center at https://www.law.go.kr/lsInfoP.do?lsiSeq=268543 (Framework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust (AI Basic Act, Act No. 20676)). This record links to it and is checked against it.
Disclaimer: informational only, not legal advice. Verify every claim against the linked official sources and consult a qualified lawyer before acting.