A Bank of Korea study released on the 7th found that after generative artificial intelligence (AI) was introduced to work, working hours fell by an average of 1.5 hours a week. But the reduction in working hours did not translate into higher productivity. Analysts said the organization structure prevented workers who saved time from being reassigned to other productive tasks.
According to the Bank of Korea report titled "Does AI adoption increase productivity? An analysis of the first three years," the average working hours of workers using AI fell by 1.5 hours a week, based on a 40-hour workweek. The findings are based on a survey of 5,512 employed people ages 15 to 64.
If workers who finish early are reassigned to other tasks, productivity is estimated to rise by about 1 percentage point. But the total volume of work handled did not increase. In other words, even when workers finish early using AI, they are not reassigned to additional tasks.
The Bank of Korea said, "This suggests that the time saved by using AI is not being reallocated to higher-productivity work," and noted, "Policy should focus less on adopting AI itself and more on converting it into productivity."
By contrast, among the self-employed, when AI cut working time by 1 percentage point, the volume of work handled increased by 1 percentage point. Because performance is directly tied to monetary rewards, they focused on using AI to boost productivity.
Meanwhile, the occupations that could sharply reduce working hours with AI were, in order, professionals, office workers, and managers. By contrast, the impact of AI was limited for service workers, skilled workers, and simple laborers. By task, AI was useful for developing educational materials, statistical analysis, model design, and software development. By contrast, AI could not play a role in work requiring physical cooperation, such as operating equipment.