2023 年 22 巻 p. 51-68
The use of artificial intelligence (AI) to foster innovation has been growing over the recent years from social and political perspectives. However, the existing research is very limited and has not well examined whether the use of AI encourages firm-level innovation. This study employs firm-level data obtained from the Japanese National Innovation Survey 2020 and quantitatively examines the relationship between the use of AI and innovation or not. The study focuses on the AI use of start-ups and small-and-medium firms, which are likely to face resource constraints for their innovation activities. The empirical results indicate the following three possibilities. First, the share of new products to total sales in 2020 (hereafter, new-product-sales ratio) are higher for firms using AI and for those using AI in a broader scope of their innovation processes. Second, the impact of the use of AI on the new-product-sales ratio is the highest for firms using AI for new product development, followed by those using AI for data collection and analysis and decision-making supports and by those using AI for the automation of operation and cost reduction. On the contrary, the impact of the use of AI on the new-product-sales ratio is not significant for firms that use AI to improve existing products. Third, start-ups that use AI for new-product development achieve a higher new-product-sales ratio than SMEs and start-ups that do not use AI. Additional supplementary analysis indicates that start-ups are more likely to train employees and to hire highly skilled employees such as post-graduates, suggesting that the advantages for start-ups that use AI for innovation arise from high-skilled human capital.