Taiwan Governance Model

説明を展開

Taiwan's AI governance is a "framework law + risk tiering + multi-agency coordination" model: (1) the AI Basic Act passed in December 2025, designating NSTC as competent authority and codifying seven international principles — sustainability and well-being, human autonomy, privacy and data governance, cybersecurity and safety, transparency and explainability, fairness and non-discrimination, accountability; (2) the Executive Yuan is mandated to set up a National AI Strategy Special Committee; (3) MODA must establish an internationally aligned risk classification framework; (4) high-risk AI applications must be labelled and labour rights safeguarded; (5) the implementation timeline is concrete — within 3 months: minors/human rights/gender impact assessment; within 6 months: risk assessment of existing government AI uses; within 12 months: government AI usage rules; within 24 months: alignment of relevant statutes. Assessment: Taiwan chose an EU-style "framework law + secondary regulations" governance path, more legally prescriptive than Singapore's "tools and codes" route, but specifics depend on the 24-month rule-making window.

参考出典

• AI Basic Act passed (Focus Taiwan) · 2025-12-23
• Taiwan passes AI Basic Act (Taiwan News) · 2025-12-23
• NSTC competent authority

同じ地域で掘り下げる