Taiwan’s Tricky AI Balancing Act

As Taiwan’s labor pool dwindles, AI is showing off how capable it is at filling in the gaps. Yet there’s a fine line between filling gaps in the workforce and widening them.

Global demand for artificial intelligence is skyrocketing, forcing countries to strike a delicate balance between the needs of industry and those of their workforce. Encouraged by upbeat predictions of the technology’s ability to increase productivity and reduce costs, more businesses are pushing to adopt AI in the workplace, even as workers grow increasingly apprehensive about the technology’s potential to replace them.

That anxiety is evident even in Taiwan, which produces most of the advanced semiconductors and other infrastructure that underpin the Age of AI. A June survey by 104 Job Bank, a Taiwanese recruitment and job search platform, found that just over a third of jobseekers aged 18-30 are highly concerned that their value to employers could decrease due to AI.

As AI becomes more deeply embedded in the workplace, the question is less about whether companies will adopt it, and more about how they will manage that adoption. Many are now calling for more stringent guidelines to safeguard systems and people. Still others say such guidelines must be flexible enough to both allow businesses to grow and ensure that workers are protected.

For Taiwan, striking that balance is increasingly important as it faces a plunging birthrate and super-aged population, where more than 20% of people are aged 65 and older. Looming labor shortages caused by this situation have companies looking to AI as a tool to maintain the high level of productivity Taiwan’s export-oriented economy is known for. But given Taiwan’s leading position in advanced technology, coupled with a highly educated workforce and a recent history of good governance, the country already has many of the tools needed to develop and implement AI policy that benefits all of society.

Rapid adoption

Recent data indicates that Taiwanese companies are already rapidly adopting AI, making the need to craft effective AI guidelines even more urgent. A Q3 survey from business and data analytics firm Dun & Bradstreet (D&B) finds that around 99% of businesses in Taiwan have launched AI-related projects. A quarter of those had already achieved operationalization, while 41% were being scaled, and another 27% remained in the planning stage.

And while jobseekers are feeling anxious about the potential negative impact of AI, employed workers appear to have a much different perspective on the technology’s growing role in the workplace. In May, talent solutions and recruitment company Robert Walters released the results of its most recent survey on AI use in countries across Asia. It found that 79% of Taiwanese employees expect a positive effect from AI on their career opportunities, 62% use AI for predictive analytics and decision support, and a staggering 96% use AI at work, the highest figure of any other country included in survey.  

D&B’s take on the use of AI by its own employees is optimistic yet measured. “Internally, we’ve built a culture that actively encourages AI adoption,” says Julian Prower, business operations EVP and transformation general manager of D&B Taiwan. He says D&B runs monthly AI use-case sharing sessions where teams across the organization discuss how they’re applying AI tools in their day-to-day work.

“The goal is to make experimentation a norm, not an exception,” Prower says. “At the same time, we take data security seriously, and that standard hasn’t changed just because AI is involved. If anything, it’s become more important.” He notes that D&B maintains strict controls on the use of AI tools, reflecting the principles it advocates for its clients.

Speaking about the need to build a coherent bridge between AI implementation and workplace regulations and expectations, Prower says: “Innovation without governance creates risk — governance without innovation creates stagnation.” He adds that, “the challenge for Taiwanese businesses is how to incorporate AI systems that are simultaneously explainable, auditable, and grounded in verified business context, so that humans remain in control of important decisions.”

Another issue both businesses and workers must contend with is the inconsistency in quality of work that AI tools produce. “While AI is expected to generate tremendous organizational value, unreliable and incomplete outputs continue to frustrate individual workers,” says William Zyzo, managing director of Z&A Knowledge Solutions and an advisor to AmCham’s Advanced Learning Lab. “Until large language models become more dependable, our prediction is that AI-related gains will remain fragmented across specific tasks, workflows, and individual workers.”

When asked whether any of Z&A’s clients had reported changes in employee attitudes toward using AI for their work, Zyzo replies that his firm hasn’t looked at the trend systematically. “However, just recently, a regional-level marketing head at a financial company who was initially an AI skeptic became quite happy with it when she found that it could generate posters for presentations she did not want to do but had to.”

Zyzo offers other examples of individual employees using AI. Some found it useful for specific tasks like cleaning up text in another language. “Other times, the same people get disappointed with it, such as when the AI doesn’t immediately give them what they want,” he says.

Training the workforce

While the majority of AI development is being driven by large multinational corporations, small and medium-sized enterprises (SMEs), which account for more than 90% of Taiwan’s industrial economy, will be the main test bed for AI use in the country’s workplaces.

To help these companies with AI adoption, Taiwan’s Ministry of Economic Affairs (MOEA) has rolled out two related but distinct programs. In February, it launched the 2026 SME Intelligence Management Efficiency Enhancement Project, which offers enterprise diagnostic services to see what business problems can be solved with AI, industry-specific training, and implementation of a tailored AI solution.

Then in July, the MOEA announced that its “Industrial Competitiveness Guidance Team” initiative, launched in October last year, had helped more than 9,200 companies adopt AI, nearly halfway toward the program’s goal of reaching 20,000 companies by 2028. Around 93% of those that have signed up are SMEs, Industrial Development Administration Director General Chiou Chyou-huey said at a July press conference in Taipei.

The initiative involves a three-pronged approach to help businesses incorporate AI in their operations. The first prong is assessing companies’ operational needs and providing a consultation valued at NT$200,000 (US$6,000), the majority of which is subsidized by the government. The second prong is AI adoption, which involves a grant for AI software and hardware deployment totaling NT$100,000, plus an additional NT$120,000 in training vouchers. Lastly, the MOEA is offering R&D grants of up to NT$5 million per AI-enabled project, as well as NT$40 million for large-scale consortium projects.

At the press conference, Chiou said that so far, the program has generated NT$32.3 billion (US$1 billion) in output value by enhancing research and development efficiency, reducing defect rates, and improving production processes

According to the 2026 Taiwan Industry AI Adoption Survey published in the official MOEA report in July, the overall AI adoption rate among Taiwan’s manufacturing sector has risen from 12.3% in 2024 to 38.5% this year. The MOEA aims to increase this figure to more than 50%, and to generate output of NT$50 billion through AI adoption and AI-driven industrial transformation by 2028.

The MOEA is not the only government ministry pushing for strategic, targeted AI adoption in the workplace. The Ministry of Education (MOE) in July introduced its Future Talent AI Competency Framework to build four core competencies among students and young professionals: using AI to learn, using it responsibly, using it to solve problems, and developing skills that AI cannot replace.

The primary vehicle through which the MOE plans to carry out that vision in universities is the Taiwan Artificial Intelligence College Alliance, which offers credit programs covering various aspects of AI industrial application.

The ultimate goal of these programs is to cultivate a pool of talent that is skilled both in their professional field and possesses practical AI capability, what Taiwan President Lai Ching-te has referred to as “two swords.” His administration aims to cultivate 500,000 such AI professionals by 2040.

But whether that goal matches Taiwan’s reality is another question. In a June 26 feature article in the Taipei Times, Taipei-based communications expert Nigel P. Daly wrote that “the problem is not only that young people lack newly emerging AI skills, but also, importantly, that schools, training programs, employers, and government agencies have not agreed on what applied AI readiness actually looks like.” He argues that Taiwan needs a more systematized approach, including mentor-based training and internships that can lead to real work outcomes.

Z&A’s Zyzo offers a different view. “I believe that the success of the MOE’s AI framework should be evaluated not only by its inspirational framing, but mainly by whether students can successfully integrate their AI proficiency with domain expertise, critical judgment, ethical reasoning, creativity, communication, collaboration, and leadership in solving real problems that Taiwan faces as a democratic society.” He further states that AI should remain a tool that serves education’s broader purpose, focusing on developing citizens who are capable of creating both value and meaning for themselves and society.

That ambition will only become more important as companies rush to make AI a focal point of their future business operations. But the window of opportunity may still be open for Taiwan to lead the way in ensuring the adoption of AI in the workplace is good for businesses, workers, and society at large.