Taiwan - AI Action Plan 2.0

Overview

The Taiwan AI Action Plan 2.0 (2023-2026) is a comprehensive national strategy approved by the Executive Yuan in June 2023 to advance Taiwan's artificial intelligence ecosystem. Building on AI Action Plan 1.0 (2018-2021), which invested NT$40 billion and achieved 48.7% corporate AI adoption, the 2.0 version targets growing Taiwan's AI industry to over NT$250 billion (US$8.16 billion) by 2026. The National Science and Technology Council (NSTC) coordinates implementation alongside the Ministry of Digital Affairs (MODA). The plan commits approximately NT$9-10 billion annually to transform Taiwan's semiconductor leadership into generative AI capabilities, establishing research centers, training programs, and regulatory frameworks that position Taiwan as a global hub for trustworthy AI innovation.

Definitions

The Action Plan establishes key terminology for Taiwan's AI strategy. Artificial Intelligence encompasses machine-based systems making predictions, recommendations, or decisions. Generative AI refers to AI systems capable of creating text, images, code, and other content based on training data. Edge AI describes AI processing performed locally on devices rather than cloud servers. AI Ecosystem refers to the integrated network of talent, research, industry, and infrastructure supporting AI development. Smart Taiwan describes the vision of AI-enhanced government services and industrial competitiveness. AI Evaluation means systematic assessment of AI products for safety, fairness, and compliance with standards.

Governance and Institutional Framework

The Executive Yuan provides cabinet-level oversight and funding authorization for the Action Plan. The National Science and Technology Council (NSTC) serves as the primary coordinating body, having been renamed from the Ministry of Science and Technology in 2022. NSTC is responsible for AI policy development, research coordination, and the draft AI Basic Act. The Ministry of Digital Affairs (MODA) was mandated to establish the AI Evaluation Centre, opened in December 2023, to assess AI products and systems. The Ministry of Economic Affairs supports industrial development and international partnerships. University AI research centers across Taiwan contribute academic expertise and talent development. The multi-ministry approach ensures comprehensive coverage of AI development from basic research through commercialization and governance.

Key Focus Areas

• Talent Development: Training AI specialists across universities, vocational programs, and industry partnerships to meet growing demand for AI expertise.
• Generative AI Development: Leveraging Taiwan's semiconductor strengths to advance large language models and content generation capabilities.
• Industry Growth: Supporting AI adoption across sectors and fostering AI startup ecosystem development.
• Semiconductor Integration: Connecting Taiwan's world-leading chip manufacturing to AI hardware development and edge computing.
• AI Evaluation Framework: Establishing the AI Evaluation Centre and certification mechanisms for trustworthy AI products.
• Research Infrastructure: Expanding computing resources, datasets, and research facilities for AI development.
• Government AI Adoption: Deploying AI solutions across public sector services for efficiency and cost savings.
• International Collaboration: Building partnerships with global technology companies and research institutions.
• Regulatory Framework: Developing the draft AI Basic Act and governance principles for responsible AI deployment.
• Edge AI Innovation: Advancing on-device AI processing capabilities for privacy and performance benefits.
• AI Ethics and Safety: Integrating ethical considerations and safety requirements into AI development practices.
• Smart City Applications: Deploying AI for urban management, transportation, and public services.

Implementation Framework

Implementation proceeds through coordinated efforts across government ministries, research institutions, and private sector partners. The Executive Yuan allocates approximately NT$9-10 billion annually through ministerial budgets, with NSTC coordinating research priorities and MODA leading regulatory development. University AI research centers receive funding for basic research and talent cultivation programs. Industry partnerships, exemplified by Google's Smart Taiwan Initiative, Microsoft's R&D center, and Synopsys' AI design center in Hsinchu, bring global expertise and investment. The AI Evaluation Centre provides certification services enabling responsible AI deployment. Free GPU computing resources support AI startups in model training, reducing development costs. Phased implementation priorities include research infrastructure expansion, talent pipeline development, industry ecosystem growth, and regulatory framework establishment. Annual reviews assess progress against targets and enable course corrections.

Monitoring and Evaluation

The Action Plan establishes key performance indicators tracked through annual assessments. Primary metrics include AI industry value (target: NT$250 billion by 2026), corporate AI adoption rates (baseline: 48.7% from Plan 1.0), trained AI specialists, and R&D investment levels. The NSTC conducts the Taiwan AI Readiness Assessment, providing regular measurement of progress against international benchmarks. Ministry-level reporting tracks implementation of specific initiatives within their mandates. University research centers report on academic outputs, talent development, and technology transfer. Industry partnerships are evaluated for investment, job creation, and innovation outcomes. The AI Evaluation Centre tracks certification volumes and product compliance. International rankings and assessments provide external validation of Taiwan's AI capabilities relative to global peers. Findings inform budget allocations and priority adjustments in subsequent planning cycles.

Penalties, Liability, and Appeals

As a strategic policy framework rather than binding legislation, the AI Action Plan 2.0 does not establish direct penalties for non-compliance. Implementation accountability flows through government ministry performance expectations and budget processes. Funding recipients must meet specified deliverables and reporting requirements. The draft AI Basic Act, developed under the Action Plan mandate, would establish more specific obligations and potential penalties upon legislative enactment. AI products failing MODA evaluation do not receive certification, limiting market access and government procurement eligibility. Existing legal frameworks including consumer protection, product liability, and sector-specific regulations apply to AI-related harms. The regulatory framework is expected to evolve as the AI Basic Act progresses through the legislative process, with more defined penalty provisions emerging in the comprehensive AI governance regime.

Relationship to Other Instruments

The AI Action Plan 2.0 provides strategic context for Taiwan's evolving AI governance framework. The draft AI Basic Act, mandated by the Action Plan and published for consultation in July 2024, would establish the primary legislative foundation for AI regulation. The AI Product and System Evaluation Centre, required by the Action Plan and opened in December 2023, operationalizes product assessment and certification. The Personal Data Protection Act (PDPA) provides foundational privacy requirements that AI systems must respect. Cyber Security Management Act requirements inform security assessment protocols. Industry-specific regulations in healthcare, finance, and telecommunications establish sector requirements for AI deployment. International frameworks including OECD AI Principles and EU AI Act provisions inform Taiwan's approach to ethical AI development and international alignment. The Action Plan connects these instruments into a coherent ecosystem supporting trustworthy AI development.

International Alignment

Taiwan's AI strategy explicitly positions the nation within global AI governance frameworks while leveraging international partnerships. Alignment with OECD AI Principles informs ethical development practices and governance approaches. The AI Evaluation Centre's criteria incorporate EU AI Act requirements, enabling Taiwanese products to demonstrate compliance with the world's most comprehensive AI regulation. US frameworks including the NIST AI Risk Management Framework inform technical assessment methodologies. Major technology company partnerships bring global best practices and investment: Google's Smart Taiwan Initiative, Microsoft's R&D center, and Synopsys' AI design center in Hsinchu demonstrate Taiwan's attractiveness for international AI investment. Participation in international forums on AI governance contributes Taiwanese perspectives on semiconductor-enabled AI development. This international orientation supports export competitiveness while enabling Taiwan to learn from global experiences in AI governance.

Implementation Timeline

Sources and References