Buying an artificial-intelligence licence is easy. Rebuilding a company so that the tool produces verifiable results is another operation.

A Reuters survey conducted by Nikkei Research shows that more than 80% of Japanese companies have not fully integrated AI. Only 16% say they use it across the organisation and, in many cases, activity still centres on basic tasks such as creating documents.

About 60% use it in limited areas. Eighteen per cent have not decided what to do and 6% are not considering adoption. The survey ran from 29 July to 6 August and received replies from 219 of 510 companies contacted.

The number does not mean Japan ignores the technology. Most companies have experimented. It reveals the gap between individual use and integration: moving from one person summarising text to a system connected to data, decisions and organisational responsibility.

AdvertisementARQUITHEAArchitecture for seeing more clearlyIdeas, buildings and tools for understanding the city through real questions.Follow @arquithea_

That second stage requires organising information scattered across files, departments and suppliers. It also means deciding who verifies an answer, which data may leave the company and what happens when a model is wrong.

Companies expect to keep spending. None said they would reduce annual AI budgets over the next year or two. Yet only 4% expect an increase above 50%, while roughly a third foresee modest or single-digit growth.

01 / 04

Gradual investment can be prudent. It avoids imposing a platform before value is proven. It can also leave scattered pilots that never change the complete process while accumulating licences, duplicated data and incompatible rules.

Architecture and design offices will recognise the pattern. Generating an image or summarising a brief offers an immediate gain. Integrating AI into cost planning, codes, BIM documentation or quality control means connecting files and accepting professional consequences.

Deep adoption is not measured by how many employees open a chatbot. It is measured by time saved, errors avoided, quality maintained and the ability to reconstruct how a decision was made.

Japan also prefers domestic investment. Forty-one per cent of companies plan to increase it by March 2027, while 50% will maintain current levels. That preference connects AI to semiconductors, robotics, employment and data sovereignty.

Public pressure says Japan must close the gap with the US and China. But accelerating adoption without changing processes may create a collection of assistants with little aggregate effect and greater dependence on external suppliers.

The fact. 16% report complete organisational deployment; 60% remain in limited use. The editorial reading is that the next advantage will not come from testing a tool first, but from redesigning work, control and knowledge so that a successful test can scale.

The question

Does slow adoption mean Japan is losing the race?

Not necessarily. Careful testing may avoid dependency and expensive mistakes. It becomes delay when trials fail to create data, owners and processes able to scale what actually works.