Risk aversion is slowing wider AI adoption at Japanese companies
Japanese businesses are using generative AI for everyday tasks such as writing and research, but caution over errors and reputational risk is reportedly slowing its adoption for mo
By BBC News
Japanese companies are being held back from making wider use of artificial intelligence by risk aversion, slow decision-making and a workplace culture with little tolerance for errors, according to business experts.
While generative AI is increasingly being used for everyday tasks such as writing, summarising and gathering information, adoption in more important business processes remains more cautious.
The issue highlights a growing challenge for companies attempting to move beyond experimenting with AI towards using the technology to change how their businesses actually operate.
Parrisa Haghirian, professor of international management at the Kyoto University of Advanced Science, told BBC News that many Japanese companies remained highly risk averse when considering broader use of artificial intelligence.
Businesses can be particularly sensitive to errors, uncertainty and the potential reputational consequences of allowing AI systems to perform more important tasks.
That matters because generative AI does not always produce predictable or completely accurate results.
For lower-risk activities, such as preparing an initial draft of a document or summarising information, an employee can review the AI’s output before it is used.
Giving the technology greater responsibility for operational or commercial decisions requires considerably more confidence in how the system behaves.
Austin Xu, co-founder of US start-up Kuse AI, also pointed to the way decisions are made inside some Japanese organisations.
He said process and consensus culture could slow the adoption of new technology, with companies sometimes preferring to leave positions unfilled rather than transfer the work to machines.
Japan presents an unusual case for AI adoption because the country also faces significant demographic pressures.
An ageing population and labour shortages create a potentially strong economic case for technologies capable of automating work and increasing employee productivity.
But the availability of the technology does not automatically mean businesses are prepared to reorganise their operations around it.
The experience highlights a wider distinction emerging as companies around the world invest in generative AI.
Using an AI assistant to draft an email, summarise a report or find information requires relatively little organisational change.
Allowing AI to make decisions, communicate autonomously with customers or control important business processes introduces questions around accuracy, accountability, security and who is responsible when something goes wrong.
Companies therefore face a balance between moving quickly enough to capture potential productivity improvements and maintaining sufficient oversight to prevent costly mistakes.
For businesses outside Japan, the experience provides a broader lesson about AI adoption.
The biggest barrier may increasingly be less about obtaining access to powerful AI models and more about deciding which responsibilities organisations are prepared to hand over to them.
As the technology becomes more capable, companies able to establish effective controls while still allowing employees to experiment with AI may gain an advantage over businesses that either adopt it without sufficient safeguards or remain too cautious to use it beyond basic tasks.