north-asia
Taiwan
legislative draft
Quick Overview
Taiwan is taking a major step toward formal governance with its proposed Basic Act on Artificial Intelligence.
Unlike neighbouring markets that rely on principles or sector rules, Taiwan’s Act introduces structured accountability and risk-based oversight.
The draft aims to promote innovation while ensuring that significant systems operate safely and transparently.
What's Changing
- The National Development Council (NDC) released the draft Basic Act on Artificial Intelligence for consultation in 2025.
- The Act introduces risk tiering for general, limited-risk, and high-risk use cases.
- Developers and deployers must meet transparency, documentation, and safety-testing obligations depending on risk class.
- Public-sector systems with material impact will require registration and impact disclosures.
- Administrative fines and corrective orders are proposed for serious non-compliance.
Who's Affected
- Technology companies creating or integrating predictive systems.
- Public-sector agencies deploying systems that influence welfare, education, or administrative decisions.
- Enterprises using automated systems in hiring, credit, insurance, and healthcare.
- Foreign vendors selling tools or platforms used in Taiwan.
Core Principles
- Accountability: Shared duties across developers and deployers.
- Transparency: Clear documentation of automated decisions.
- Fairness: Systems must not create discriminatory outcomes.
- Data protection: Alignment with Taiwan’s Personal Data Protection Act.
- Risk proportion: Oversight increases with potential impact.
What It Means for Business
- Companies operating in Taiwan should prepare early by mapping their system portfolio and identifying risk levels.
- Vendor contracts should reflect shared responsibility for fairness, testing, and documentation.
- Pilot audits and internal governance structures will position organisations well ahead of formal enforcement.
- Adopting international standards (ISO/IEC, OECD) can simplify compliance.
What to Watch Next
- Parliamentary debate and amendments to the draft Basic Act.
- Creation of a national oversight authority or coordination office.
- Pilot audits in public services before full enforcement.
- Coordination with Japan, Singapore, and the EU on risk-tiering standards.
| Aspect | Taiwan | Japan | South Korea |
|---|---|---|---|
| Approach Type | Draft legislation | Principles | Rights-based |
| Legal Strength | Pending | Voluntary | Moderate |
| Focus Areas | Accountability, risk levels | Fairness, safety | Privacy, fairness |
| Lead Bodies | National Development Council | METI, Cabinet Office | MSIT, PIPC |
Local Resources
Related coverage on AIinASIA explores how these policies affect businesses, platforms, and adoption across the region. View AI regulation coverage
This overview is provided for general informational purposes only and does not constitute legal advice. Regulatory frameworks may evolve, and readers should consult official government sources or legal counsel where appropriate.
Latest Comments (4)
It's good to see Taiwan is *finally* drafting an AI Act, but is "proportional oversight" really enough? In Korea, we've seen how quickly tech can outpace regulation. Folks often focus on innovation, yet perhaps the *true* challenge is how to enforce accountability without stifling the development of novel applications. Just my two cents.
This is quite interesting, Taiwan really stepping up to the plate. While balancing innovation and accountability is certainly the goal we all want, I wonder how they intend to practically enforce these "proportional oversight" measures, especially when dealing with the lightning-fast development cycle of AI. Will there be a dedicated regulatory body with tech experts, or will it fall under existing ministries? That’s the real million-dollar question, innit? Seems like a proper challenge to get that right without stifling future breakthroughs.
This is promising news coming out of Taiwan. It's smart to get ahead on AI regulations, especially with all the rapid developments we're seeing globally. My main query, though, is how they plan to actually *enforce* these accountability measures, particularly when so much AI development is borderless. Will this act have real teeth against international companies operating in Taiwan, or will it be more of a guideline? That's the million-dollar question, isn't it? Hope it sets a good precedent for other nations, including our own.
Interesting, but I wonder if they're forgetting AI's potential for grassroots uplift in developing nations. Accountability can sometimes stifle innovation, right?
Leave a Comment