Korea AI Basic Act — trustworthy AI obligations

Note: this explainer is informational and framework references are never a compliance guarantee — align your program with your legal counsel.

South Korea's AI Basic Act — officially the Basic Act on the Development of Artificial Intelligence and the Establishment of Foundation for Trustworthiness — was passed on 27 December 2024 and took effect on 22 January 2026 (with a grace period on penalties). It is one of the first comprehensive national AI laws outside the EU. Official notice: Korea MSIT — Basic Act on AI; US government overview: trade.gov — South Korea AI Basic Act.

Who it applies to

The Act reaches AI businesses that develop or provide AI systems in South Korea, and — through a domestic representative requirement — foreign operators above a threshold whose services reach Korean users. If your workforce uses AI tools that process Korean personal data or serve the Korean market, it is in scope for your governance planning.

The core obligations

The Act frames trustworthy AI around a few operating duties:

• High-impact AI — systems used in areas that materially affect life, safety, or fundamental rights carry heightened risk-management, documentation, and user-protection duties.
• Transparency — users must be informed when they interact with AI, and generative AI outputs must be identifiable as AI-generated.
• Accountability — providers are expected to keep records that demonstrate risk management and responsible operation.

The common thread is knowing where AI is used and being able to show how it is governed.

How Shield maps to it

Shield gives you the operating layer those duties assume:

• Discovery shows which AI tools — including generative-AI tools — are actually in use across the workforce, so oversight starts from reality: review discovered tools.
• Policies and enforcement modes govern what data reaches high-impact tools and how usage is controlled: enforcement modes.
• The audit log keeps a tamper-resistant, exportable record that evidences responsible operation: audit log and export.

Applying the Korea AI Basic Act bundle from the preset library turns these into framework-labelled policies in minutes.