To keep pace with the rapid evolution of artificial intelligence (AI) and address the challenges it brings, the Executive Yuan approved the draft bill for a basic law on AI on August 28, 2025. Aimed at promoting AI innovation while safeguarding people's rights and interests, the proposed law outlines seven core principles to guide AI research, development and application: sustainability, human autonomy, privacy protection and data governance, safety, transparency and explainability, fairness, and accountability. The draft legislation serves to anchor Taiwan's key priorities for AI development, helping to foster an environment conducive to AI research and application, thereby positioning Taiwan as a global leader in the field. In addition, it emphasizes the protection of national sovereignty and cultural values, and the creation of a smarter, safer and more sustainable digital society.
Main points
■ Innovation through collaboration: Promote AI research, development, application and infrastructure through grants, commissioned projects, investments, rewards, and policy measures such as tax and financial incentives. Encourage international cooperation and facilitate the cross-border exchange and utilization of talent, technology and infrastructure. Establish mechanisms for data openness and sharing to enhance the availability of AI-related data.
■ Talent cultivation: Foster AI education across schools, industries, public institutions and society to strengthen AI literacy.
■ Research flexibility: Build and enhance innovative testbeds for AI research, development and application services while easing restrictions on academic research to accelerate the real-world deployment of AI technologies.
■ Risk management: Advance the adoption of a risk classification framework aligned with international standards, and develop mechanisms for accountability, remedy, compensation and insurance in high-risk AI applications.
■ Assessment and verification: Establish standards and create mechanisms for verification, traceability and accountability, based on risk management principles, to enhance the verifiability and human oversight of AI applications. Utilize assessment and verification tools to help ensure legal compliance of AI applications, and clarify accountability for high-risk AI applications.
■ Safeguarding of rights and interests: Provide employment counseling for workers displaced by AI. Ensure that personal data is collected, processed or used only when necessary during AI research, development and application. Increase the quality and quantity of data used in AI to preserve the nation's cultural diversity and protect intellectual property rights.