Analyzing the direct role of governmental organizations in artificial intelligence innovation

Abstract

Artificial intelligence (AI) has rapidly emerged as a transformative technology with the potential to revolutionize numerous industries and applications. While government organizations actively support the AI innovation ecosystem through funding and policy making, their active and direct participation through patenting has not been well studied. Here, we analyzes the intramural patenting activity of government organizations and compares it to that of non-governmental organizations in the field of AI. By analyzing the representative terms in the AI patent abstracts and patent matching using machine-learning-based document embedding, we found that governmental organizations more focus on public benefit and national-level interests, rather than commercialization, which is a main focus of non-governmental organizations. Moreover, our regression results reveal that the AI patents by governmental organizations are cited by more diverse areas than non-governmental organizations, which shows their wider impacts on future innovation. Our findings contribute to the literature on the role of government in fostering innovation in the field of AI and have implications for policy makers and stakeholders involved in AI R&D funding and commercialization.