What question did this study set out to answer?

This work examines Korea's AI Framework Act through the principles of business and human rights.

July 31, 2026Open Access

‘Light-Touch Rights’ in AI Governance: A Business and Human Rights Analysis of South Korea’s AI Framework Act

Key Points

This work examines Korea's AI Framework Act through the principles of business and human rights.
Analyzed the AI Framework Act in relation to the UN Guiding Principles on Business and Human Rights.
Evaluated the Act's enforcement and commitment measures across its three pillars.
Assessed the implications of the Act's design for AI governance and human rights.
The Act lacks provisions for unacceptable-risk AI, leading to weak enforcement (Pillar I).
Human rights due diligence is fragmented across voluntary obligations (Pillar II).
Limited pathways for remediation exist for individuals harmed by AI systems (Pillar III).

Abstract

Abstract In January 2025, the Republic of Korea became the first country in Asia to enact comprehensive national artificial intelligence (AI) legislation. This piece analyses Korea’s AI Framework Act, which entered into force in January 2026, through the lens of the United Nations Guiding Principles on Business and Human Rights (UNGPs). It argues that the Act establishes a ‘light-touch rights’ model: its industrial promotion measures are backed by binding commitments, while its human rights protections rely largely on procedural duties and best-effort provisions. Assessed against the three pillars of the UNGPs, the Act omits a prohibited category for unacceptable-risk AI and provides weak enforcement (Pillar I), disperses human rights due diligence across voluntary and procedural obligations (Pillar II), and offers no meaningful remedial pathways for those harmed by AI systems (Pillar III). Without reform, Korea risks exporting a template of ‘rights without remedies’ across the region.

Ask AI

Helpful

Bookmark

View Full Paper