Japanese corporations are increasingly implementing systems that link employees' artificial intelligence (AI) skills to bonuses, performance evaluations, and promotions. This shift encourages employees to utilize AI and directly ties their actual work performance to compensation and evaluations.
The Nihon Keizai Shimbun reported on July 12 that Honda, a major Japanese automotive company, has established a program that awards employees recognized for their AI skills with monthly bonuses of up to 150,000 yen (approximately $1,400). This program targets about 45,000 employees across Japan, including those in research facilities.
Honda evaluates employees' AI skills on a three-tier scale, providing bonuses in addition to their base salary. The assessment is based on written tests, interviews, and actual work performance. Employees who achieve concrete results, such as developing programs using AI, can attain higher ratings. As of July 2026, there are 280 certified employees, with plans to increase this number to 1,000 within a few years.
Currently, around 10 employees receive the top-tier bonus of 150,000 yen. This highest rating is awarded to those who can propose AI utilization strategies tailored to various departments, including research and development, production, and management. Honda expects these employees to disseminate AI usage knowledge among their colleagues and promote its adoption within the company.
More companies are also incorporating AI performance into employee evaluations. FamilyMart has required approximately 4,500 employees to include AI utilization plans in their work objectives since April, with the degree of achievement and improvements reflected in performance evaluations. All Nippon Airways (ANA) plans to incorporate AI-driven work improvements into evaluations and salaries starting in the 2024 fiscal year. For instance, a mechanic developed a system that uses AI image analysis for visual inspections of aircraft engines. ANA evaluates AI performance separately from regular work and sets different target levels based on job grades, allowing lower-graded employees to receive high evaluations for significant achievements.
Mitsubishi Corporation plans to make passing the 'G Certification' exam, administered by the Japan Deep Learning Association, a requirement for promotion to managerial positions starting in the 2027 fiscal year. Sumitomo Corporation evaluates all employees' AI skills on a six-tier scale based on certification and work performance.
The push for bonuses and promotions linked to AI skills stems from concerns that AI utilization is limited to certain departments. A survey by the Japanese recruitment firm Persol Career in February found that only 20% of over 500 manufacturing and telecommunications companies used AI company-wide. A report by Microsoft in May indicated that Japan ranked 48th globally in generative AI adoption, with a rate of 22.5%. In comparison, South Korea ranked 16th and the United States 21st, while the United Arab Emirates and Singapore topped the list.
However, rewarding employees solely for high AI usage can lead to negative consequences. Amazon recently halted its ranking system for employee AI tool usage after some employees began using AI agents for unnecessary tasks. This shift turned the initiative into a competition for increased usage rather than meaningful application. Salesforce has introduced metrics to track how much employees' AI usage contributes to their work performance.
Differences in AI utilization opportunities and training conditions across job types complicate the establishment of evaluation criteria. Shohei Gino, a researcher at Recruit Works Institute, noted that environments for learning AI technology are often unequal, and requiring uniform utilization from all employees could create perceptions of unfairness.
As of now, few Japanese companies have integrated AI into their performance evaluations. A survey by Nihon Keizai Shimbun in March found that less than 10% of 143 major companies had adopted related systems. The challenge for these companies is not how much employees use AI, but how effectively they improve their work through AI.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.