Why Japanese Firms Are Slow to Use AI
Monday, 2026/08/17211 words3 minutes592 reads
Despite facing labour shortages, an ageing population, and persistently low productivity, Japan has been notably cautious in embracing artificial intelligence in the workplace. According to an OECD report, only 8.4% of Japanese workers were using AI as part of their jobs at the end of 2025—a stark contrast to separate estimates of 50% in the United States and 32% in the United Kingdom.
Experts interviewed by the BBC attribute this hesitancy to deeply ingrained organisational characteristics. Japanese companies typically operate with consensus-driven decision-making processes and maintain an exceptionally low tolerance for errors, particularly in client-facing operations. Consequently, generative AI is predominantly deployed for lower-risk activities such as writing, summarising, and information gathering, rather than for core operations or strategic decision-making.
Structural barriers compound these cultural factors. Approximately 60% of corporate computer systems are over two decades old, and many healthcare facilities have yet to fully digitalise patient records. Furthermore, Japan faces a projected shortfall of nearly 800,000 IT professionals by 2030. While the government's 2025 AI Promotion Act adopts a light-touch regulatory approach, and some large firms are beginning to prioritise AI literacy in recruitment, autonomous multi-step AI systems remain uncommon. The central question persists: can Japan leverage AI for productivity gains while preserving its emphasis on reliability and risk management?
