Firms’ influence on the evolution of published knowledge when a science-related technology emerges: the case of artificial intelligence

Firms with the assets complementary to Artificial Intelligence (AI) have actively conducted AI research and selectively published their results since AI has resurged around 2006. Focusing on the recent AI development, we investigate how and to what extent firms’ deep engagement in the publication of emerging science-related technology can influence the evolution of published knowledge. Using bibliometric analyses applied to the papers in major AI conferences and journals, we find that papers with at least one author affiliated to a firm, and particularly papers with only firm-affiliated author(s), have had higher influence on the formation of published knowledge trajectory than other papers. In addition, papers from firm and non-firm (university and public research institution) collaborations show higher novelty and conventionality than other papers. These findings deepen our understanding of the role of firms in the evolution of emerging science-related technology.