Bank of Korea Report: Generative AI Adoption Hits 66.5%, But Link Between Time Savings and Output Remains Weak
A new Bank of Korea survey shows significant growth in workplace generative AI use and time savings in South Korea, but highlights limited depth in core tasks and weak correlation with actual output improvements.
The Employment Research Team at the Bank of Korea's Survey Department released a report titled 'AI Diffusion and Productivity Effects: More Widespread, More Efficient, But Still Shallow' on November 17. The data indicates that the proportion of South Korean workers who have used generative AI for work increased from 51.8% last year to 66.5% this year, with regular users rising from 17.1% to 30.6%. This adoption rate is significantly higher than the average of 33.7% observed in the US and six European countries.
As AI tool usage has spread, reported work time savings have also increased substantially. The average time-saving rate for employees who used AI last week rose from 3.8% last year to 9.2% this year, equivalent to saving about 3.7 hours in a standard 40-hour work week. Managers, professionals, and administrative staff reported savings rates around 11%. Paid users reported higher time savings (11.4%-11.6%) compared to free users (8.3%), though the study notes this may reflect higher demand and reliance in certain roles rather than solely the effect of paid services.
Despite the increase in time savings, the link between these efficiency gains and actual output remains weak. The report shows a correlation coefficient of only 0.08 between time-saving rates and increases in work volume, and 0.09 with quality improvement rates. This suggests that employees saving more time are not necessarily processing more work or significantly improving quality. Furthermore, in 'core tasks' essential to job functions, actual AI application levels fall far short of theoretically possible substitution rates, indicating AI is currently used more for auxiliary or non-core activities.
The study also found that age and education remain key factors influencing AI adoption. Workers aged 18-39 and those with graduate degrees showed significantly higher usage rates than those over 50 or with lower educational attainment. The gender gap has largely disappeared, with female usage (66.6%) now nearly equal to male usage (66.5%). The Bank of Korea recommends focusing future efforts on converting AI-driven time savings into tangible productivity growth and evaluating comprehensive benefits including quality improvements and worker well-being.