Generative AI and Employment in Japan ― How to ensure young people develop skills ―

Generative AI (hereinafter, "GenAI"), the prime example of which is ChatGPT, has become capable of performing tasks such as drafting documents, organizing data, and handling routine responses in place of humans. These tasks have traditionally been entry-level duties that enable young people who have just entered the workforce to learn the basics of work. In the U.S., the employment of young workers for entry-level tasks is visibly declining due to the impact of GenAI. Meanwhile, employment for veteran workers remains stable, and it has been noted that GenAI is bringing about a "seniority-biased technological change."

In Japan, labor shortages caused by the declining birthrate and aging population, as well as employment practices such as emphasis on a sense of belonging to an organization (below, "membership-based employment") and the simultaneous hiring of new graduates, are serving as cushions, making it unlikely that unemployment among young workers will surge as it has in the U.S. But what is concerning in Japan is not the possibility of a reduction in the number of jobs, but rather the problem of young people getting older without being able to develop vocational skills.

Japan's membership-based employment has developed human resources through a process in which workers begin with simple tasks, grow through on-the-job training (OJT), and are gradually entrusted with more important work. If GenAI takes over the starting point of this process, the development mechanism is in danger of collapsing from within. There is a risk that time will pass without young people being able to acquire the skills that would normally be picked up gradually through OJT, i.e., practical training in the workplace, and a "generation that was able to get jobs, but could not learn about work" may emerge. This problem will only become apparent five to ten years from now, when the cohort that joined companies during the proliferation of GenAI become managers responsible for business decisions. Measures must be taken now before it becomes too late.

There is a notion that companies should take responsibility for training young people, but that cannot solve this problem. First, in a society where the declining birthrate means that young people are scarce, training them increases their value, making it easy for them to switch employers. As a result, companies are reluctant to invest in training young people. Second, even if the quality of experience declines, it is unlikely that this will be reflected in numerical indicators, delaying recognition of the problem. Third, amid intense competition with rivals, investment in training tends to be pushed back. The upshot is that society as a whole is in danger of falling into a vicious cycle in which investment in training young people keeps decreasing.

Policies to prevent such an "experience gap" from developing can be grouped into three categories: (1) expand fiscal and institutional support for companies that invest in training young people in parallel with the introduction of GenAI, and establish pre-employment training opportunities at universities and colleges of technology; (2) require disclosure of investment in human capital so that companies do not stop "investing in people" in favor of short-term profits, thereby allowing market discipline to function; and (3) build data foundations, including national statistics, to continuously monitor and visualize the impact of GenAI on entry-level tasks and experience accumulation for young people across society.

Japan's so-called "Employment Ice Age" was a period in which employment opportunities themselves disappeared. But what is coming next is a problem of employment quality, namely a "generation that was able to get jobs, but was unable to acquire skills." To avert such a situation, it is necessary for the public and private sectors to work together and take action now.

PDF link(PDF:660KB)