Does Artificial Intelligence Basic Act require Risk Assessment?

Taiwan • enforcing

Yes — 1 provision

Requirements at a glance

This regulation imposes 6 specific requirements for Risk Assessment across 1 provision:

Government AI Governance Framework #

Obligation:
Risk Assessment
enforcing
Effective:
Jan 14, 2026
Risk tier:
all
Scope:
Government bodies. The Ministry of Digital Affairs must promote an internationally interoperable AI risk taxonomy and assessment framework (Art. 16(1)); sectoral competent authorities must then establish risk-based management regulations and assist their industries in producing guidelines and codes of conduct (Art. 16(2)); the government must conduct risk assessments before using AI to perform duties or provide services (Art. 19). The Act imposes no direct compliance duty on private-sector developers or deployers — obligations reach industry only once a sectoral regulator issues rules under Article 16(2)
sleeperupcoming
A framework act, not a compliance statute. Every operative article directs the state: fund AI development (Arts. 9-10), open government data (Art. 13), protect labour rights (Art. 15), clarify high-risk liability and establish relief or insurance mechanisms (Art. 17), and review all conflicting law within two years (Art. 18). Private-sector obligations arrive later and indirectly, through whatever sectoral regulators issue under Article 16(2) — which is what makes the two-year Article 18 deadline of 2028-01-14 the date to watch. **Open scope question:** on a strict reading of exclusion principle E4 (wrong audience), this Act may belong in `data/exclusions.md` rather than as a tracked instrument, since it creates no private-sector duty. It is retained for now because it is the enabling frame for every future Taiwanese AI rule; revisit when the first Article 16(2) sectoral regulations appear.

Requirements

RequirementDetails
Risk taxonomyThe Ministry of Digital Affairs must promote an AI risk taxonomy and assessment framework interoperable with international standards, and assist sectoral authorities in establishing risk-based management regulations (Article 16(1))
Sectoral rulemakingSectoral competent authorities must establish risk-based management regulations following that taxonomy and assist their industries in formulating guidelines and codes of conduct (Article 16(2))
High-risk liabilityFor high-risk AI applications, the government must clarify liability attribution and conditions and establish relief, compensation, or insurance mechanisms; this does not reach pre-application R&D unless tested in a real-world environment or used to provide products or services (Article 17)
Conforming legal reviewWithin two years of the effective date, the government must complete the enactment, amendment, or repeal of non-conforming laws, regulations, and administrative measures (Article 18)
Government use assessmentWhen using AI to perform duties or provide services, the government must conduct risk assessments, plan response measures, and establish usage guidelines or internal control mechanisms (Article 19)
Data protection by designSectoral authorities, consulting the personal data protection authority, must avoid unnecessary collection, processing, or use of personal data in AI R&D and application and promote data protection by design and by default (Article 14)

Penalties

ViolationFine
NoneThe Act contains no penalty provision. It directs government action and carries no sanction against private parties
View full regulation View obligation Obligation matrix