Is it true that "AI cannot be developed without Japan"? — From Ajinomoto to SoftBank, the invisible leadership of Japanese companies in the AI era

In the news surrounding generative AI, names of American companies like OpenAI, Google, Microsoft, and NVIDIA appear almost daily. Massive foundational models, GPUs, and cloud services are spoken of as the main players in the AI revolution, and there is a growing perception that Japan has fallen behind in this new digital competition.

However, when we broaden our perspective to include the physical supply chain that powers AI, a different picture emerges.

Materials that enable high-performance semiconductors, memory for storing data, and precision components and equipment that support manufacturing processes—behind the glamorous AI services lies the irreplaceable technology of Japanese companies. While Japan may not fully control the "brains" of AI, it supports parts of the body and nerves that make those brains into real products.

The view presented by overseas management company Comgest, that "Japanese companies are indispensable to global AI production," although somewhat strongly worded, offers a perspective for reevaluating Japanese companies.

The unexpected role of a food company in supporting AI chips

Ajinomoto is a symbolic presence.

To the general consumer, Ajinomoto is a food company dealing in seasonings and frozen foods. However, the "Ajinomoto Build-up Film" or ABF, developed by the company group, is an important insulating material used in the package substrates of high-performance semiconductors.

The circuits inside semiconductor chips are miniaturized to the nanometer level. On the other hand, the wiring on the printed circuit board side, where the chip is mounted, is much thicker. ABF enables the formation of multilayer substrates that connect the two, serving as the foundation for transmitting fine electrical signals externally.

For AI GPUs and server processors, as processing power increases, the demands for wiring density, heat resistance, and substrate flatness also rise. Therefore, as the demand for AI semiconductors grows, the importance of materials for advanced packaging will also increase.

Interestingly, ABF was born from the extension of amino acid research and resin technology cultivated in the food business. The result of continuing seemingly unrelated research over a long period is a business that supports global computing.

This is an important case when considering the strengths of Japanese companies. Japanese companies have lagged behind American companies in developing internet services that dominate the market in a short period. However, in fields where they have improved specific materials and processes over decades and continuously refined quality in collaboration with client companies, they remain strong.

The advantage Japan holds in the AI era did not suddenly appear. Past research and development, coordination with the field, and accumulated failures are beginning to transform into strategic assets today.

Kioxia targets the huge demand for "AI inference"

In AI infrastructure, not only computing semiconductors are in the spotlight, but also the memory that stores generated data and learning information is indispensable.

Kioxia Holdings is one of the few companies that supply NAND flash memory on a global scale. Although it is a business susceptible to fluctuations in demand for smartphones and PCs, the expansion of AI data centers is creating new growth opportunities.

Particularly important is the demand for "inference."

In the learning stage of developing AI models, a large number of GPUs and high-speed memory are needed. However, as AI becomes widespread in society, inference processing, such as answering user questions, searching company data, and generating images or videos, will occur routinely.

As the number of inferences increases, access to stored data and reading of model information increase, raising the required capacity and speed of storage. When AI usage shifts from the experimental stage to full-scale operation, flash memory companies may greatly benefit.

However, the memory business is subject to significant market fluctuations. If companies simultaneously increase production capacity in anticipation of demand expansion, oversupply could lead to price declines. Strong AI demand alone does not guarantee sustained high profitability.

What Japanese companies need is not just to increase production volume, but to expand areas less prone to price competition, such as high-performance products, energy-saving products, and customer-specific designs.

Is SoftBank an "outlier" or "pioneer" among Japanese companies?

In contrast to companies supplying materials and components, SoftBank Group is investing large sums of money into the core of the AI industry.

The company is the controlling shareholder of semiconductor design company Arm and has significantly expanded its cumulative investment in OpenAI. If the additional investment plan announced in February 2026 is completed, the cumulative investment in OpenAI will be $64.6 billion, with an ownership ratio of about 13%.

SoftBank is attempting to view AI models, semiconductor design, robots, and data centers as a single economic zone. The concept is not just to hold a part of AI services but to broadly cover everything from the computing infrastructure to the user terminals.

This approach is significantly different from traditional Japanese companies, which excel in cautious investment and gradual improvement. SoftBank's stance of boldly predicting future market size and investing huge sums while leveraging borrowings and equity holdings is, for better or worse, atypical among Japanese companies.

The rise in OpenAI's corporate value has significantly boosted SoftBank's profits. However, there are also strong concerns about the concentration of investment in specific companies, large-scale fundraising, and the profitability of AI data centers.

If AI changes the world economy as expected, SoftBank could achieve historic investment results. However, if competition between models intensifies, usage prices fall, or there is excess capacity, the investment burden could become heavy.

Here, two choices that Japanese companies face in the AI era are condensed: whether to remain as component suppliers with limited value shares or to take significant risks and advance into upstream industries or platforms.

Changes in Japanese companies that cannot be explained by AI alone

The reevaluation of Japanese companies is not occurring only in AI-related stocks.

Chugai Pharmaceutical has accumulated unique drug discovery technologies focused on cancer and immune diseases and is expanding its research results to the global market through its partnership with Roche. Its strength lies in not relying solely on the domestic market but expanding intellectual property and license income globally.

Fast Retailing's Uniqlo business is also continuing to grow in North America and Europe. This is the result of maintaining Japanese-style meticulous quality control while adapting product planning, inventory management, store operations, and brand communication to each region.

What both companies have in common is that they are not moving towards a shrinking equilibrium due to Japan's declining population but are taking the capabilities honed domestically to the global market.

In the past, Japanese companies grew by exporting large quantities of finished products. Today's leading companies provide value overseas in different forms, such as materials, intellectual property, pharmaceuticals, brands, data, and operational know-how.

When considering the international competitiveness of Japanese companies, the era of looking only at the number of exports of electrical products and automobiles is coming to an end.

"Japanese funds" starting to buy Japanese stocks

The original article also cites the increase in domestic stock investment by Japanese institutional investors as a tailwind.

However, this needs to be viewed carefully.

The Government Pension Investment Fund (GPIF) significantly raised its basic allocation to domestic stocks in 2014. Subsequently, it has been set at 25% for the period from fiscal 2020, and the new basic portfolio from fiscal 2025 also maintains 25%.

In other words, GPIF has not significantly raised its domestic stock ratio recently. While it is true that the allocation to domestic stocks has increased in the long term, it is not appropriate to explain the recent rise in Japanese stocks solely by GPIF's new purchases.

Rather, what should be noted as a current change is that the Tokyo Stock Exchange is demanding listed companies to manage with an awareness of capital costs and stock prices.

There is increasing pressure on companies that hold excessive cash, continue to have unprofitable businesses, or lack explanations to shareholders to improve. It is not just about share buybacks and dividend increases, but also about business sales, reduction of policy-held stocks, and investment in growth areas, questioning the very use of capital.

If performance growth due to AI demand and corporate governance reforms progress simultaneously, the evaluation of Japanese stocks may shift from a temporary undervaluation correction to continuous growth expectations.

"Ikigai" should not be beautified as endurance

A distinctive feature of Comgest's claim is that it explains the long-term strength of Japanese companies with the word "ikigai."

Japanese companies have continued their business while responding to resource shortages, natural disasters, limited land, and intense international competition. The view that a corporate culture emphasizing not only short-term profits but also social roles and pride in technology has supported long-term research and development and quality improvement has a certain persuasiveness.

However, the Japanese word "ikigai" does not originally mean only dedication to the company or enduring hardships. It includes personal and voluntary values such as the joy of daily life, relationships with others, a sense of being needed, and the fulfillment of engaging in what one loves.

Simplifying it into a corporate narrative of "Japanese people are strong because they endure hardships" is dangerous.

In Japan, long working hours, peer pressure, seniority, and slow decision-making have long been issues. If employee endurance is rebranded as competitiveness, it could become an excuse to preserve old labor practices.

The ikigai needed in the AI era is not for humans to work even harder to compete with AI. It is about shifting routine tasks and document creation to AI, creating a state where humans can spend time on more proactive work such as empathy with customers, choosing research themes, design, negotiation, and judgment on the ground.

If companies demand a sense of purpose from their employees, they must also provide the authority, time, and room for failure for employees to challenge themselves at their own discretion.

"Surprise" and "pride" spreading on social media

In public SNS posts about ABF, the surprise that "a seasoning company is supporting AI semiconductors" is prominent.

Especially on overseas platforms like X and Reddit, the unexpectedness of Ajinomoto being a food company yet having a dominant presence in semiconductor materials is repeatedly discussed. The story that "a company seen in supermarkets is also part of the data center supply chain" is easy to understand and tends to spread as a tale symbolizing the technological prowess of Japanese companies.

There is also an evaluation that Japanese companies are poor at advertising but actually hold important parts of the global market. There is strong anticipation for companies and products not known to the general consumer to suddenly come into the spotlight due to the AI boom.

These reactions can be an opportunity to regain confidence in Japanese manufacturing. Even if Japan is lagging behind the US and China in the AI model development competition, there is a view that Japan has a different way of fighting.

Concerns about supply concentration

On the other hand, users knowledgeable about investment and semiconductors have pointed out the possibility that ABF's supply capacity could become a bottleneck for the entire AI industry.

Materials with a high dependency on one or a few companies become a strong competitive advantage for the supplier but also a risk for global customers. If a factory accident, natural disaster, raw material shortage, or quality issue occurs, it could affect a wide range of product supplies.

If production capacity is rapidly increased, there may be excess facilities when future demand decreases. If increased cautiously, supply shortages may occur when demand surges. This difficult investment decision is also being discussed on social media.

While there are voices evaluating the management stance of suppressing short-term price increases despite having an overwhelming share, there are also opinions that "Japanese companies with strong technology should earn more appropriate profits and reinvest in research and development."

This is also a challenge that Japanese companies have faced for many years. Simply supplying high-quality parts at low prices makes it difficult to earn as much profit as companies that hold the final product or platform.

Having technological strength and recovering sufficient economic value from that technology are separate issues.

Reactions to SoftBank are divided

Reactions on social media to SoftBank's investment in OpenAI are even more divided than those to Ajinomoto.

The positive side appreciates the significance of Japanese capital being deeply involved in a global AI company. This is because there is a possibility that Japanese companies can benefit from the growth of the AI platform itself, not just as component suppliers.

The negative side is concerned about the size of the investment, reliance on borrowing, and concentration on OpenAI. Citing past investments like WeWork, there are many voices warning that SoftBank's bold investments also come with the possibility of significant losses.

This approval and disapproval also reflect differing evaluations of the entire AI market.

For those who believe AI will bring about an industrial transformation greater than the internet or smartphones, SoftBank's investment seems rational. For those who think current valuations and capital investments are too far ahead of actual profits, it looks like a huge bubble.

Which is correct has not been settled at this point.

Discomfort with the word "IKIGAI"

Discussions linking AI and ikigai are met with both empathy and discomfort.

Positive reactions consider that as AI replaces human jobs, the focus shifts from "what to streamline" to "why we work." The Japanese way of emphasizing not just skills but intrinsic motivation and connection with society is argued to be useful in building organizations in the AI era.

On the other hand, there is criticism that "IKIGAI" is simplified overseas and has become a convenient term for self-improvement and corporate training.

If companies demand employees to "work with ikigai" without improving treatment, discretion, and working hours, it shifts the responsibility onto individuals. Management that demands a sense of mission from remaining employees while advancing personnel reductions through AI introduction is the opposite of ikigai.

Ikigai is not a slogan given by the company to employees. It is something the individual discovers for themselves, and the company's role is likely to create an environment that does not take it away.

Can Japan's AI utilization break away from being "inward-looking"?

AI adoption by Japanese companies is progressing, but the main purposes tend to be efficiency improvements such as minute-taking, document summarization, internal searches, and inquiry responses.