News
Major insurer Tokio Marine Holdings has announced a partnership with OpenAI to jointly develop AI agents for use in sales and customer support. The initiative is expected to extend beyond operational efficiency into areas such as product design and risk assessment.
Source: Reuters
Meanwhile, Toyota Motor is building its experimental “Woven City” at the foot of Mt. Fuji in Susono, Shizuoka. The project aims to integrate AI, autonomous driving, and robotics into a “city of the future.” With Japan’s aging and shrinking population in mind, it seeks to tackle healthcare, welfare, and transportation challenges—drawing global attention.
Source: Toyota Motor Official Announcement
Background
Tokio Marine’s AI agent project and Toyota’s Woven City vision symbolize how Japanese conglomerates are finally beginning to incorporate AI into real-world applications. SoftBank is also working with OpenAI to build a domestic AI data center, while experimental adoption is expanding across manufacturing and municipal governments.
However, in the United States and China, AI has already reached commercial-level use in finance, retail, and government. Japan lags far behind in adoption, investment, and talent development. Surveys show that about 70% of professionals in the U.S. and over 80% in China use AI at work, while Japan remains below 30%.
In this sense, these recent moves can be seen as Japan’s first steps toward closing its “AI adoption gap.” To explore this issue further, let’s look at how it was debated on Reddit in the thread:
“OPINION: Has Japan already capitulated in the great global AI race?”
Overseas Reactions
Here are selected comments from that thread:
The United States and China dominate AI due to vast data resources, global linguistic reach, and massive R&D investment. Japan’s AI usage is only 26.7%, compared to 69% in the U.S. and 81% in China. Japanese companies are focusing AI usage on cost-cutting rather than innovation. With limited language data, weak funding, and low adoption, Japan ranks low in Stanford’s AI Vibrancy ranking. Its digital trade deficit could balloon to ¥45 trillion by 2035.
Probably because adoption of deep AI (or agentic AI) is difficult in a country with a risk-averse culture with a historical reliance on human capital and manual processes. No early-stage funding scene to fuel innovation and small companies needed to create transition. Literally none.
Don’t forget stupidity.
Yes? Young people can barely use computers as it is. Big corporations and old Showa men killed Japan’s young IT whiz kids in the 1990s and 2000s with strict copyright laws and by bowing to American pressure. The mass media didn’t help either, portraying computers as “otaku” culture or harmful to society. This is Japan’s own doing.
What was the American pressure? I worked in the music industry in the late 2000s, and Japanese companies were incredibly slow in adopting digital content technologies, even from American firms.
When I was in Japan in 2016, people were still using ガラケー flip phones and even PSP UMDs for music. Now in the 2020s, I don’t see much of either, but adoption of new tech is still slow. Even trying to get a job as a 3D artist using Blender was difficult because everyone stuck with Maya. Crazy that I could earn more as a Blender artist in Slovakia than in Japan.
We had new hires who had literally never touched a PC in their smartphone-centric lives.
Computer literacy is abysmal for Gen Z and Alpha all over the globe. Smartphones are to blame.
So much bullshit in so many words.
This is it. Dead on the nail.
Despite repeated claims of AI boosting productivity, nothing of true value has emerged in years. Maybe that’s not such a bad thing?
I don’t know your industry, but in programming, Cursor has already become quite common. Of course, it’s just a supporting tool, not a replacement.
I’m a programmer. Many colleagues claim AI has boosted productivity tenfold, but in reality they don’t work less or produce more. Something feels off.
Also a programmer, and same here. Never used it, never will.
I am a programmer. Cursor is basically useless in real projects except for boilerplate code. Stop glorifying AI coding—it can’t replace programmers and never will. Sure, AI can handle boring repetitive tasks, but anything requiring practical experience is out of reach. Automation and AI sound good, but in reality Japan has tried and failed multiple times—like automated convenience stores. They exist, but hardly anyone uses them, since Japanese conbini are not just for groceries but also for bills, tickets, postal, and delivery services. Yes, Toyota has successful automation in its factories, but overall, today’s AI hype feels like the California Gold Rush—Nvidia and AI founders profit while investors are fooled. Slap “AI” on a product and it sells.
You guys are reading way too much into this (though you’re not wrong). The simpler reason: it’s easier to build a large language model when more than a billion people use the language. Only English and Chinese qualify.
Bingo.
Lol, that assumes AI is mostly for language-intensive tasks. In reality, AI productivity gains have come mostly in programming, where human languages matter little.
Yes, it’s a large language model. And “human” languages? C# is a language too. Feed an LLM enough code, it’ll excel at it. Of course, even state-of-the-art models are still weak compared to average programmers. But the fact remains—if most of the high-quality data is in English or Chinese, those training the LLMs have an edge.
You got fooled by the AI meme—aka bubble.
Yes, few Japanese companies can build systems from scratch. But many are good at piggybacking on trends, creating “AI products” that aren’t really good but can still find buyers.
For example, DeNA (owner of the Yokohama Baystars baseball team) claims to be going “all-in for AI.” I expect their collaboration projects to flop.
Almost everything software-related is doomed here. Entertainment software is the only exception. Japan prioritizes hardware but has lost in many industries there too. When that last edge disappears, Japan’s tech sector will be in serious trouble.
This has been the case for years. Read Stanford’s 2023 and 2024 AI reports (is 2025 out yet?). Japan’s lag is no joke. I’ve written about this in academic papers. In some ways, it’s good for fields like education—gives them time to see pros and cons abroad. But surveys I’ve run with students over the past three years show something sad: the overwhelming majority of computer science and art students in Japan don’t think AI will affect their careers.
At the same time, Japanese students I spoke with—from middle to university—almost all said they use ChatGPT or other AI tools for homework.
I saw some statistics: among the world’s top 100 IT companies, the only Japanese firm was Rakuten—and only because of domestic business. Globally, they’ve failed. That’s why Japan lags in AI too.
Many of Japan’s major service websites still look like they’re from the 1990s… Buying Shinkansen tickets or booking JAL flights feels like a retro experience.
Overall, opinions pointed to cultural, financial, and historical reasons behind Japan’s slow AI adoption. Many noted the lack of PC literacy among youth and the cautious stance toward new technology. Yet some suggested this “delay” might be a blessing in disguise for education or social balance. What most agreed on, however, is that Japan remains significantly behind in the global AI race.
Analysis: Why Japan Fell Behind—and Where It Might Still Excel
Structural Reasons for Japan’s Late Start
- Risk-averse corporate culture and reliance on manual labor
- Weak early-stage funding for startups
- Limited language data and poor digital literacy in education
Unexpected Benefits of Being a Latecomer
- Opportunity to observe U.S. and China’s pitfalls (privacy, ethics, job disruption)
- Time to craft safer, more sustainable AI adoption, especially in education and governance
SMEs and Regional Gaps
AI adoption remains centered on large corporations. SMEs and rural governments lag due to cost and talent shortages, risking a widening digital divide. Shared AI infrastructure and public support will be vital.
AI, Demographics, and Employment
With labor shortages driven by demographic decline, AI could fill gaps in routine tasks. But this also risks displacing middle-class clerical workers. Balancing these impacts will be a major policy challenge.
Education and AI Literacy
Young Japanese often enter the workforce with little PC experience, exacerbating inequality. Without improved AI literacy education, society risks a split between “AI users” and “the left behind.”
Japan’s Experimentalism vs. China’s Pragmatism
Toyota’s Woven City represents Japan’s experimental, phased approach to AI, robotics, and autonomous systems. But China has leapfrogged into real-world deployments (self-driving taxis, AI in public services), surprising even Toyota. Japan’s slower “proof-first” model ensures safety and trust, but puts it at a disadvantage in speed.
Geopolitical Dimension
The U.S. and China place AI at the heart of defense and national strategy. Japan, by contrast, has focused narrowly on business applications and “peaceful uses.” Without more active participation in AI governance and security, it risks being sidelined.
Japan’s Possible Strengths
- Manufacturing automation and precision engineering
- Entertainment (anime, gaming, creative AI)
- Services and elderly care solutions aligned with demographic needs
Conclusion: Will Japan Be a Follower or a Model?
Japan has fallen significantly behind the U.S. and China in AI adoption, due to structural, cultural, and educational issues. Yet as a latecomer, it also has unique opportunities: to learn from others’ mistakes, to leverage demographic realities like labor shortages, and to showcase trusted, sustainable AI models.
Whether Japan remains a follower or forges its own unique model will depend on bridging the SME divide, reforming education, and engaging in international AI governance. If successful, Japan could still emerge as a leader—not in speed, but in credibility and innovation tailored to its own challenges.
That’s all for now—see you again in the next article.
References
- Tokio Marine Holdings announces business partnership with OpenAI (Reuters)
- Toyota Woven City Officially Launches (Toyota Global Newsroom)
- Toyota opens experimental “Woven City” site near Mt. Fuji (Reuters)
- Stanford University – AI Index Report 2024 (Full PDF)
- Pew Research Center – Survey on AI and Workplace Use (2024)
- MIT Technology Review – Latest Case Studies of AI Implementation in China
