South Korea AI law took effect today as the Basic Act on the Development of Artificial Intelligence and Establishment of Trust began its implementation, creating a comprehensive national framework to govern artificial intelligence technologies.
The law, which was passed by the National Assembly in December 2024, consolidates multiple AI-related bills into a unified legal structure and aims to balance innovation with responsibility in AI development and deployment.
The AI Basic Act was approved through successive committee votes and a plenary session before presidential authorization, setting January 22, 2026, as the formal enforcement date.
The legislation represents a broad effort to establish national governance, promote ethical AI use, and protect citizens from risks posed by advanced AI systems.
National Framework and Governance
The act creates a legal foundation for long-term AI policy planning. Under its provisions, the Minister of Science and ICT may develop a new AI Master Plan every three years to strengthen the nation’s competitiveness in AI.
It also formalizes a National AI Committee, chaired by the President, to guide policy, investment, infrastructure, and regulatory actions across government agencies.
In addition, the law establishes an AI Safety Institute tasked with protecting public safety and property from potential AI-related risks.
These governance structures are intended to centralize oversight and coordinate interagency efforts across the technology landscape.
Risk-Based Rules for AI Use
A central element of the legislation is its risk-based approach to AI systems.
The law lays out specific categories such as high-impact AI, defined to include systems that significantly affect human life, safety, or fundamental rights.
Organizations deploying these systems must undertake risk assessments, establish safety measures, maintain human oversight, and protect users.
The act also addresses generative AI systems, requiring businesses to notify users when outputs are AI-generated, especially when such content is difficult to distinguish from real-world data.
These transparency provisions aim to curb deception and build consumer trust in AI-driven content.
Scope and Compliance
The law has extraterritorial reach, applying to AI activities outside South Korea that affect its domestic market.
Under this provision, non-domestic AI businesses that meet certain thresholds must appoint a domestic representative to ensure compliance with Korean regulations and to liaise with authorities on obligations under the act.
Enforcement and Draft Rules
While the act’s main provisions are now in force, the Ministry of Science and ICT has released a draft enforcement decree detailing implementation elements including safety standards, transparency criteria, and definitions for high-performance AI models.
The draft is open for public review and reflects input from industry and the national strategy committee.
The government has signaled a grace period into 2026 for businesses to adjust to compliance requirements before enforcement actions begin.
The decree also includes definitions such as cumulative computational thresholds for categorizing high-performance AI models.
