日本語 Get in touch

AI implementation reports for companies in Japan

Insights

Written from Japanese government primary sources only. Every number here has a citation you can check yourself.

25 articles

  1. 01 AI skills

    The 38.8% who never learned how

    58.8% of people in Japan have now used generative AI, up from 26.7% a year earlier. Of those who have not, 38.8% say the reason is that they do not know how. That is the highest share of the four countries surveyed, and it is a training problem, not an attitude problem.     3 min read
  2. 02 International

    Your next customer does not read Japanese

    Japan had 3,768,977 foreign residents at the end of 2024, a record, up 10.5% in one year. In the year to October 2024 the Japanese national population fell by 898,000. If your product, your website and your support only work in Japanese, you are opting out of the only part of the market that is growing.     2 min read
  3. 03 Automation

    898,000 fewer people, and the work did not shrink

    Japan's population fell by 550,000 in the year to October 2024. The Japanese national population fell by 898,000. The working age share is 59.6% and the share aged 65 and over is 29.3%, a record. Automation in Japan is no longer a productivity project. It is how the work gets done at all.     2 min read
  4. 04 Governance

    If your staff use AI, the AI Guidelines for Business are about you

    Japan's AI Guidelines for Business reached version 1.2 on 31 March 2026. They name three roles, developer, provider and user, and ten principles. Most companies assume they apply to vendors. The moment an employee uses a model in their work, the company is an AI user under the guidelines, and the user has obligations too.     3 min read
  5. 05 Privacy

    The two sentences from the privacy regulator your staff have not read

    In June 2023 Japan's Personal Information Protection Commission told every business handling personal data two things about generative AI. Check that any personal data in a prompt is within the stated purpose of use. And before entering personal data, check how the provider handles it, because if it is used for anything beyond answering you, consent is needed. Three years on, many companies still have no written rule that says either.     3 min read
  6. 06 AI adoption

    Why your AI pilot never became a product

    27.0% of Japanese companies report no organisational effort around generative AI. The same question puts the United States at 1.4%. What separates them is not the tooling.     3 min read
  7. 07 AI adoption

    Japan caught up on adoption and not on effect

    86.4% of Japanese companies use generative AI. The United States is at 90.9%. The adoption gap has essentially closed. What remains is the gap in what changed afterwards.     2 min read
  8. 08 AI usage

    Japan treats AI as a dictionary, the US treats it as a colleague

    Asked what generative AI is most like, the second most common Japanese answer was "dictionary or encyclopaedia". In the US and China it was "friend". That difference produces a difference in results.     3 min read
  9. 09 Engineering

    Only 7.0% of Japanese companies use AI to write code

    Japan reports 7.0% usage of generative AI for code generation. The United States is at 35.4%, China 33.5%, Germany 25.0%. This is where the gap in software speed comes from.     3 min read
  10. 10 Talent

    3.8% of Japanese companies say their DX talent is good enough

    In IPA's survey, 3.8% of Japanese firms said they have adequate DX talent by quality. The United States answered 52.9%. Meanwhile 80% report a shortage and 19.4% do no hiring at all.     3 min read
  11. 11 Design management

    0.8% of Japanese manufacturers credit design for their success

    Surveyed across 25,000 manufacturers with 100 or more staff, 38.3% credited quality and function for success in the domestic market. Design got 0.8%. For the US market, design got 0.0%.     2 min read
  12. 12 Productivity

    Japanese manufacturing went from first in the OECD to twentieth

    In 2000 Japan ranked first in the OECD for manufacturing labour productivity. The 2025 survey puts it 20th of 35. Measured per hour across the whole economy, Japan is 28th of 38.     2 min read
  13. 13 Policy

    A Cabinet decision now asks for a division of roles between people and AI

    Japan's Basic Plan for Artificial Intelligence, adopted 23 December 2025, states that the obstacle is implementation rather than research. Its fourth pillar is effectively a design brief.     2 min read
  14. 14 Policy

    Japan's AI Act has no penalties and still creates a duty for users

    The AI Promotion Act (Act No. 53 of 2025) took full effect on 1 September 2025 with no penalty provisions. Article 7 places an efforts duty not only on developers but on companies that merely use AI.     2 min read
  15. 15 Accessibility

    Reasonable accommodation became mandatory in 2024 with no standard attached

    Since 1 April 2024, providing reasonable accommodation has been a legal duty for private businesses in Japan. No standard states what a website has to meet in order to satisfy it.     3 min read
  16. 16 Accessibility

    The European Accessibility Act applies to Japanese exporters

    The EU's European Accessibility Act has applied since 28 June 2025. If you sell products or services into the EU, being based outside the EU does not exempt you.     2 min read
  17. 17 Accessibility

    10.0% of Japanese over 65 bank online, against 94.2% in Sweden

    In a Cabinet Office international comparison, the share of over 65s using the internet for banking was 10.0% in Japan, 73.9% in the US, 31.7% in Germany and 94.2% in Sweden. The most aged country uses it least.     3 min read
  18. 18 Design

    People over 75 now outnumber those aged 65 to 74

    As of 1 October 2025, 17.3% of Japan's population is 75 or older, against 12.1% aged 65 to 74. What the phrase "designing for older users" refers to has quietly changed.     2 min read
  19. 19 Design systems

    Japan's Digital Agency design system is MIT licensed

    The tokens and reference components of the Digital Agency Design System are published on GitHub under the MIT licence. They were built for government services and private companies are permitted to use them.     2 min read
  20. 20 Management

    Japan subsidises both AI adoption and the retraining around it

    The IT introduction subsidy became the Digitalisation and AI Introduction Subsidy in 2026, with a ceiling of ¥45 million and a rate reaching three quarters. Training costs are covered separately at up to 75%.     2 min read
  21. 21 Productivity

    Developers were 19% slower with AI and felt 20% faster

    In a randomised controlled trial with experienced developers, working with AI tools took 19% longer. The same participants reported that AI had made them 20% faster.     2 min read
  22. 22 Talent

    3.54% annual productivity growth would erase Japan's IT talent shortage

    METI projects an IT talent shortfall of up to 790,000 people by 2030. The same report contains one line stating that supply and demand balance at 3.54% annual productivity growth. The baseline assumption is 0.7%.     2 min read
  23. 23 DX

    Legacy systems are not the main obstacle to DX

    In IPA's survey, fewer than 20% of companies in Japan, the United States and Germany call legacy a major obstacle. Modernisation activity has plateaued. Something else is doing the blocking.     2 min read
  24. 24 Marketing

    Being cited by AI now matters more than ranking in search

    Individual generative AI use in Japan has reached 58.8%. The starting point for finding information is shifting from search engines to AI. That changes who reads your company website.     2 min read
  25. 25 Process design

    Training does not fix a badly designed internal system

    Customer facing products get improvement budgets. Internal systems do not. The cost of a bad internal system still gets paid, every day, in employee time that nobody counts.     2 min read

Recognise the problem

We design, build and hand over the layer that turns a stalled AI pilot into something people actually use. Tell us where you are stuck.

Get in touch