The Ministry of Economic Affairs now aims to help more than 20,000 companies in Taiwan adopt artificial intelligence by 2028, raising an earlier target of 14,000 set last October. The expansion reflects both strong early demand and a deliberate push to lower barriers for small and medium-sized enterprises that face high costs and technical hurdles.
Since the industrial competitiveness task force launched in October, 9,249 firms have joined the program, 93 percent of them SMEs. Service-sector businesses account for 60 percent of participants, led by wholesale, retail and food service, while manufacturers - mainly in metal products and machinery - make up the remaining 40 percent.
A three-stage subsidy model
Industrial Development Administration Director-General Chiou Chyou-huey described a staged process designed to reduce risk for companies with limited resources. The first phase provides diagnostic consulting: the government covers NT$190,000 (US$5,904) of the cost and the company pays NT$10,000. Next, firms receive a NT$100,000 grant for AI hardware and software deployment, plus NT$120,000 in training vouchers.
The third stage funds deeper adoption through research-and-development grants of up to NT$5 million per project, or NT$40 million for consortium projects. For government teams designing similar incentive structures, this tiered approach offers a practical template - one that AI for Policy Makers training often examines as a case study in balancing public investment with private accountability.
Output value and adoption rates climb
Participating companies have generated about NT$32.3 billion in output value so far by improving production processes, cutting defect rates and boosting R&D efficiency. AI adoption among these firms has reached 38.5 percent, and the ministry wants to push that past 50 percent while generating NT$50 billion in output value by 2028.
AI as a workforce tool, not a replacement
Chiou pushed back on fears that AI adoption will trigger mass layoffs. "AI is mainly being used to replace repetitive tasks, and those positions are difficult to fill, rather than replacing workers," he said. He added that companies adopting AI with the expectation of cutting jobs would "regret it."
The government encourages businesses to use AI to boost productivity, preserve experienced workers' knowledge and increase employees' value rather than reduce headcounts. For yacht builder Kha Shing Enterprise, a 50-year-old company, retiring veteran workers had made knowledge transfer a critical challenge. Company president Howard Gung said AI now lets young workers quickly access information for design, materials management and pricing, while reducing rework and quotation errors.
Why this matters for government professionals
The program's emphasis on SMEs and traditional industries - not just tech firms - shows how public agencies can structure AI adoption incentives to reach sectors that typically lag. The three-stage model, with escalating grants tied to demonstrated progress, provides a blueprint for subsidy design that limits waste and builds capability step by step. Chiou's workforce message also carries a clear policy implication: positioning AI as a response to labor shortages, rather than a cause of unemployment, can defuse political resistance and align economic goals with worker retention. For officials shaping similar programs, relevant AI for Government Courses explore how to design and measure these kinds of interventions.
Your membership also unlocks:
