Newsletter
Executive Yuan Releases its Version of Draft Artificial Intelligence Basic Act
Executive Yuan Releases its Version of Draft Artificial Intelligence Basic Act
By Ken-Ying Tseng/Kai-Yao Cheng
While the Legislative Yuan has not reached a consensus on the various versions of the Artificial Intelligence Basic Act bills proposed by legislators from different political parties before the current session ended on August 29, 2025, the Executive Yuan adopted its own version of the "Artificial Intelligence Basic Act" draft (hereinafter referred to as the "Draft Basic Act") on August 28, 2025, and decided to submit it to the Legislative Yuan for review. Originally, it was the National Science and Technology Committee that was designated with the responsibility to propose and draft an Artificial Intelligence Basic Act, but the responsibility was transferred to the Ministry of Digital Affairs (the "MODA") earlier this year and this version was drafted by MODA. Below are the key points of the Draft Basic Act:
I. Core Definitions and Principles of AI
The Draft Basic Act expressly defines AI, emphasizing its autonomous capabilities to distinguish it from traditional software systems and it further sets forth seven AI principles: sustainability, human autonomy, privacy protection and data governance, security, transparency and explainability, fairness, and accountability.
II. Encouraging Innovation
The first of the two main principles of the Draft Basic Act is to encourage the innovative development of AI and to regard it as an important national policy, which includes requiring the government to take the following measures:
1. Promote AI Development and Provide Incentives: The government shall properly allocate relevant resources and offer support, commissions, investments, rewards, guidance, or fiscal incentives such as tax and financial benefits to AI-related industries.
2. Establish an Innovative Experimental Environment for AI RD and Applications: The government may refer to the regulatory sandbox system to provide controlled testing environments for AI developers and users.
3. Facilitate Public-Private Collaboration: The government shall encourage cooperation between the public and private sectors to jointly advance AI research and applications.
4. Promote International AI Cooperation And Exchange: The government shall also strive to promote international exchanges of AI talent, technology, and equipment.
5. Continuously Promote AI Talent Cultivation and Public AI Literacy Education: The government shall actively promote AI talent development across all levels of education, industry, society, and public agencies, while further enhancing AI literacy of the general public.
6. Establish Mechanisms for Data Openness, Sharing, and Reuse: To ensure the AI industry can access high-quality data for training, the government shall establish mechanisms to facilitate data openness, sharing, and reuse.
III. Safeguarding Human Rights
The second of the two main principles of the Draft Basic Act is an emphasis on the protection of human rights in the era of AI. Specifically, this includes requiring the government to take the following actions:
1. Avoid Illegal Use of AI: The government should prevent AI applications from infringing upon fundamental human rights, disrupting social order, spreading misinformation, or threatening national security, etc. and should develop assessment and verification tools accordingly.
2. Promote a Risk Classification Framework: The government should refer to international standards to promote an AI risk classification framework and establish corresponding regulations based on the level of risk.
3. Establish AI Assessment and Verification Tools: The government should develop and implement AI assessment and verification mechanisms to ensure AI systems are verifiable and human-controllable.
4. Liability and Remedies for High-Risk AI: For high-risk AI applications (excluding those still in the research and development phase), clear liability attribution, remedies, compensation, and insurance mechanisms should be established.
5. Protect Labor Rights: In response to the impact of AI development on the employment market, the government should strive to protect labor rights and provide necessary transition counseling and reemployment assistance for workers who lose their jobs as a result.
6. Personal Data Protection: The personal data protection authority should ensure the legality and security of the collection, processing, and use of personal data during AI research, development, and application.
7. Maintain Multiculturalism: The government should work to improve the quality and quantity of data used in AI to ensure that training and output results uphold the nation’s multicultural values and intellectual property rights.
After the Draft Basic Act is enacted and enters into force, (i) the government shall review and adjust its functions, operations, and regulations under its jurisdiction in accordance with the provisions of the Draft Basic Act, and (ii) before the relevant regulations have completed the required legislative procedures, the central competent authority shall consult with MODA on the interpretation and application of AI-related regulations.
The Draft Basic Act proposed by the Executive Yuan will be subject to further review by the Legislative Yuan. Our Digital Economy, TMT and Data Protection Practice Group will continue to closely monitor the developments. Should you have any questions regarding the relevant regulations, please do not hesitate to contact our team.