Policy Status
Policy status
In force
Effective date
January 2026
Applies to
Both
Regulatory impact
High
north-asia
South Korea
binding law
Quick Overview
- South Korea enacted the AI Basic Act (인공지능 기본법) on 26 December 2024, making it the first comprehensive standalone AI law in Asia.
- The law was promulgated on 21 January 2025 and took full effect on 22 January 2026, establishing a risk-based governance framework for AI development and deployment nationwide.
- The Act creates the National AI Committee under the President, mandates impact assessments for high-risk AI systems, and enshrines transparency and accountability obligations for both public and private sector AI operators.
- South Korea's AI Basic Act complements the existing Personal Information Protection Act (PIPA) and the Data 3 Act, creating one of the most comprehensive AI regulatory ecosystems in the Asia-Pacific region.
What's Changing
- The AI Basic Act introduces a risk-based classification system distinguishing between general AI and high-impact AI, with stricter obligations for systems affecting safety, fundamental rights, and critical infrastructure.
- A new National AI Committee chaired by the President will coordinate national AI strategy, set ethical guidelines, and oversee cross-ministerial regulatory alignment.
- Mandatory impact assessments are now required for high-risk AI applications in areas including healthcare, criminal justice, employment screening, and autonomous transportation.
- AI developers and operators must provide transparency notices to users, disclose the use of AI-generated content (including deepfakes), and maintain records of algorithmic decision-making processes.
- The government is investing heavily in AI safety infrastructure, including a dedicated AI Safety Research Institute and sector-specific regulatory sandboxes.
- The Act also reinforces support for AI industry growth through R&D funding, talent development programmes, and international cooperation frameworks — reflecting South Korea's strategy to balance innovation with regulation.
Who's Affected
- AI developers and service providers operating in South Korea must comply with transparency, accountability, and risk assessment requirements under the new Act.
- Government agencies deploying AI in public services — including welfare distribution, immigration screening, and law enforcement — face mandatory impact assessment obligations.
- Financial institutions, healthcare providers, and education platforms using AI for automated decision-making must meet enhanced disclosure and human oversight standards.
- Foreign technology companies offering AI services in South Korea will need to ensure compliance with the Act's transparency and data governance provisions.
- SMEs and startups benefit from regulatory sandbox provisions and government-backed AI development support programmes designed to reduce compliance burdens.
- Citizens and consumers gain new rights to explanation and redress when subject to consequential AI-driven decisions.
Core Principles
- Human dignity and rights protection: AI systems must respect fundamental rights and shall not be used in ways that undermine human dignity or democratic values.
- Transparency and explainability: AI operators must provide meaningful information about how AI systems function, including the logic behind automated decisions affecting individuals.
- Safety and reliability: High-risk AI systems must undergo rigorous testing and maintain ongoing monitoring to ensure they operate safely and as intended.
- Fairness and non-discrimination: AI systems must be designed and deployed to prevent algorithmic bias, with specific provisions against discrimination based on gender, age, disability, or regional origin.
- Accountability: Clear lines of responsibility are established for AI outcomes, with operators required to designate AI ethics officers and maintain incident response protocols.
- Innovation and balanced regulation: The Act explicitly aims to foster AI industry competitiveness while establishing necessary safeguards — embodying South Korea's philosophy of proportionate, risk-based governance.
What It Means for Business
- Companies deploying high-risk AI must conduct and document AI impact assessments before deployment, covering potential effects on safety, rights, and societal impact.
- Organisations must appoint AI ethics officers and establish internal governance frameworks aligned with the National AI Ethics Standards issued by the National AI Committee.
- Businesses using AI for consumer-facing decisions (credit scoring, hiring, insurance underwriting) must provide users with explanations of AI-driven outcomes and mechanisms for human review.
- AI-generated content — particularly deepfakes and synthetic media — must be clearly labelled, with penalties for non-compliance that apply to both creators and distributors.
- Regulatory sandboxes allow companies to test innovative AI applications under supervised conditions, with streamlined approvals for qualifying projects — a significant advantage for startups and R&D-focused firms.
- South Korea's membership in global AI governance forums (OECD, GPAI, Hiroshima AI Process) means compliance with the AI Basic Act positions businesses well for interoperability with emerging international standards.
What to Watch Next
- The National AI Committee is expected to issue detailed implementation guidelines and sector-specific standards throughout 2026, which will shape how the Act is applied in practice across industries.
- South Korea's AI Safety Research Institute is scheduled to begin operations in mid-2026, serving as a centre for AI risk assessment, incident investigation, and international research collaboration.
- Amendments to the Personal Information Protection Act (PIPA) are under discussion to better align data governance rules with the AI Basic Act's requirements for training data transparency.
- The government plans to expand regulatory sandbox programmes and announce additional AI investment incentives as part of the Digital Platform Government Initiative.
- International regulatory alignment efforts — particularly with the EU AI Act and Japan's forthcoming AI governance framework — will influence how South Korea calibrates enforcement and cross-border data flows.
- The upcoming presidential policy review in late 2026 may introduce further refinements to high-risk AI classification criteria based on early implementation experience and industry feedback.
How It Compares
← Scroll to see full table →
| Aspect | South Korea | Japan | China |
|---|---|---|---|
| Approach Type | Rights-based | Principles | Regulatory |
| Legal Strength | Moderate | Voluntary | Binding |
| Focus Areas | Privacy, accountability, fairness | Safety, fairness | Security, content |
| Lead Bodies | MSIT, PIPC | METI, Cabinet Office | CAC, MIIT |
Local Resources
Last editorial review: March 2026
Related coverage on AIinASIA explores how these policies affect businesses, platforms, and adoption across the region. View AI regulation coverage
This overview is provided for general informational purposes only and does not constitute legal advice. Regulatory frameworks may evolve, and readers should consult official government sources or legal counsel where appropriate.
