South Korea: Enactment of AI law

Why should I read this?

South Korea’s first comprehensive artificial intelligence law (the Framework Act on the Development of Artificial Intelligence and the Creation of a Foundation for Trust (Act)) took effect on January 22, 2026. This places South Korea among the first countries to introduce a national AI statute, ahead of several other regimes, including the EU AI Act. Through this law, South Korea appears to be seeking to promote rapid AI development while safeguarding safety, human rights and public trust.

The Act adopts a flexible, principles-led approach to AI regulation. It sets national standards for responsible AI development without imposing uniform rules on all AI systems. Rather than regulating every type of AI in the same way, it focuses on uses most likely to affect safety, livelihoods or fundamental rights. The aim is to reduce serious risks while still allowing innovation to continue in lower risk areas.

The Act also favors good governance over rigid technical requirements. It encourages businesses to manage AI risks through measures such as ethics committees, impact assessments and certification or verification schemes. While regulators are given enforcement powers and can impose penalties, these apply only to a limited number of clearly defined obligations. This reflects an effort to address the most serious risks without over-regulating the sector.

What are the key takeaways?

Key elements of the Act include:

• the Act establishes a National AI Committee to coordinate AI policy across government, supported by an AI Policy Center and an AI Safety Institute responsible for risk analysis and safety technologies
• the government is expected to create and regularly update a long-term AI Master Plan, linking AI development with ethics and infrastructure as part of the country’s broader national strategy.
• the Act provides a statutory basis for government funding for AI R&D, data centers, SME support, talent attraction and the promotion of AI adoption across different industries
• the Act does not regulate all AI systems in the same way. It focuses on high‑impact AI and GenAI, especially when used in sensitive areas like critical infrastructure, healthcare and transport. The scope can be expanded if new high-risk uses emerge
• developers and operators of covered AI systems are subject to specific responsibilities. They must notify users when AI is used, label AI-generated content where appropriate, ensure human oversight, manage risks, keep records and assess possible impacts on fundamental rights
• the Act promotes national AI Ethics Principles, the establishment of internal ethics committees and voluntary verification or certification schemes. Public procurement is expected to favor AI systems that have been certified or subjected to impact assessments
• authorities may request information, carry out investigations and order corrective measures. Fines of up to 30 million South Korean won can be imposed for certain failures, such as not notifying users of AI use or failing to comply with corrective orders

How does South Korea’s approach differ from other AI regulatory approaches?

South Korea has adopted a broad, joined-up approach to AI regulation. The Act combines national strategy, industrial policy and basic trust and safety rules in a single law, aiming to promote AI development while setting minimum safeguards. This contrasts with the EU AI Act, which focuses mainly on product safety, detailed technical requirements and the imposition of very high fines.

While both regimes are risk based, South Korea places greater reliance on general principles such as transparency, risk management and accountability, alongside government support and voluntary certification mechanisms. Many technical details have been left to future regulations, giving the framework greater flexibility over time.

Unlike the US, which does not have a single national AI law, South Korea provides a unified nationwide framework. Compared with jurisdictions such as Japan, Singapore and the UK, the Act goes beyond purely voluntary guidance while still relying heavily on self-regulation. Its ultimate impact will depend on future presidential decrees and regulatory guidance.