Japan’s struggle with an ageing population, labour shortages and persistently weak productivity would appear to make the country an ideal candidate for rapid adoption of artificial intelligence (AI). Yet despite growing recognition among businesses and policymakers that AI could help address some of these challenges, its use in Japanese workplaces remains considerably lower than in several other advanced economies.
Recent figures highlight the gap. According to an Organisation for Economic Co-operation and Development (OECD) report published late last year, only 8.4 per cent of workers in Japan use AI in their jobs. That compares with around 50 per cent in the United States and 32 per cent in the United Kingdom. Singapore has reported even higher workplace adoption, with 56 per cent of employees said to be using AI several times a week.
Experts say Japan’s cautious corporate culture is one of the major reasons for the slow uptake. Austin Xu, co-founder of US-based AI start-up Kuse AI, which recently established operations in Japan, believes Japanese companies tend to be highly conservative and reluctant to accept the possibility of mistakes associated with AI. In organisations where decisions are based heavily on consensus and processes are tightly controlled, introducing a technology that can produce errors can be difficult.
According to Xu, some Japanese employers may prefer to leave a position vacant rather than allow AI to perform tasks where mistakes could affect customers or the company’s reputation. This contrasts sharply with the approach seen in the US, where companies are increasingly willing to experiment with AI tools, allowing them to perform tasks and correcting problems as they emerge.
Parrisa Haghirian, a professor of international management at Kyoto University of Advanced Science, also believes Japan’s broader corporate culture makes technological change difficult. She says Japanese companies are particularly sensitive to uncertainty, mistakes and reputational damage. As a result, generative AI is largely being restricted to relatively low-risk activities such as drafting text, summarising documents and gathering information, rather than being integrated into important operational or decision-making processes.
The healthcare sector illustrates the problem particularly clearly. Some Japanese hospitals are still struggling to fully digitise patient records, leaving large volumes of information in paper form. One hospital employee described the continued reliance on paper documentation as being comparable to the “Stone Age”, highlighting the scale of the digital transformation challenge facing parts of the sector.
The Japanese government has recognised the problem and has begun taking steps to encourage greater AI adoption. Parliament passed an AI Promotion Act last year, with the government favouring a relatively light regulatory approach intended to encourage companies to invest in and develop AI technologies. Tokyo has also set an ambitious goal of making Japan one of the most supportive countries in the world for the development and use of AI.
Officials see AI as an important tool for improving productivity, an area in which Japan continues to lag behind other G7 economies. The Ministry of Finance has pointed to a sharp increase in the number of businesses using AI, claiming that adoption has risen from around 11 per cent five years ago to 75 per cent today.
However, analysts caution that such figures can be misleading. While a large proportion of companies may have introduced some form of AI technology, only a small share of employees may actually be using it, and even those users may be applying AI only for limited tasks.
Another major obstacle is Japan’s shortage of workers with advanced digital skills. Yasushi Ogasawara, a professor at Meiji University specialising in Japanese society and technology, argues that the country has relatively low digital literacy despite the popularity of consumer technology such as smartphones. A separate projection has estimated that Japan could face a shortage of nearly 800,000 IT professionals by 2030.
The problem is compounded by ageing computer infrastructure. Around 60 per cent of Japan’s business computer systems are reportedly more than two decades old, making it harder for companies to introduce modern AI-based systems and integrate them into existing operations.
Employment concerns are another factor. Ogasawara says Japanese policymakers and companies remain strongly committed to maintaining employment, making them cautious about technologies that could potentially replace workers. While the government is promoting AI, reskilling and digital education, shifting large numbers of employees into technology-oriented roles remains difficult.
Nevertheless, there are signs of change, particularly among larger companies. Some Japanese employers are beginning to include AI literacy among the qualities they seek when recruiting university graduates. Major trading company Kanematsu, for example, has recruited young employees who already use AI extensively for workplace tasks.
One recent graduate, Uta Yamaguchi, said she frequently uses AI to draft emails, summarise complex documents and process meeting recordings. However, she noted that sophisticated AI systems capable of independently managing several stages of a task remain rare in Japanese workplaces, even though her company has begun introducing such technology.
For Japan, the challenge is therefore not simply whether businesses will adopt AI, but how deeply the technology will be incorporated into everyday work. With an ageing population, a shrinking workforce and a pressing need to raise productivity, delaying the transition could carry significant economic costs.
Experts argue that Japan’s preference for gradual and low-risk reform may need to change if the country is to fully benefit from AI. While avoiding disruption may protect existing systems and jobs in the short term, it could also prevent businesses from achieving the productivity gains needed to remain competitive. For a country facing some of the developed world’s most acute demographic and labour challenges, the ability to move from cautious experimentation to meaningful AI adoption may prove increasingly important.
