Taiwan AI Basic Act
Taiwan's framework AI law (20 articles). It sets principles (sustainability, human autonomy, privacy and data governance, safety, transparency, fairness, accountability), names the National Science and Technology Council as lead authority, has the Ministry of Digital Affairs build a risk classification framework, and gives government two years to bring laws in line.
At a glance
In force
- Jurisdiction
- Taiwan
- Type
- Law
- Official title
- 人工智慧基本法 (Artificial Intelligence Basic Act)
What this means for you
If you build AI
- No direct penalties yet: sector regulators are to write risk-based rules on top of a national AI risk classification framework from the Ministry of Digital Affairs.
- R&D before real-world use is exempt from the high-risk liability rules, unless tested in real environments or offered as products or services.
If you deploy AI
- Government is to define liability, remedies, compensation or insurance for high-risk AI applications; watch for sector rules.
- Government must review its laws and fix conflicts within two years of the Act taking effect.
If you use AI
- Government agencies using AI must assess risks and set usage rules or internal controls.
Key dates
- Jan 14, 2026PastPromulgated by presidential order and in forceSource
Status history
In the Drops
No Drop has covered this law yet.