In a new PLoS ONE paper, patents are analyzed to map technological innovation dynamics in artificial intelligence (AI).
The paper, Mapping technological innovation dynamics in artificial intelligence domains: Evidence from a global patent analysis, is authored by Na Liu, Philip Shapira, Xiaoxu Yue, and Jiancheng Guan. It introduces the AI domain, examines the strengths and weaknesses of prior methods of tracking AI patents, advances a new AI patent search approach.
Using this search approach, the authors undertake a global landscape analysis of AI patenting data over two decades (1991–2020). They find that activity in AI patenting is currently in a phase of rapid growth, driven significantly (although not exclusively) by a recent increase of AI patenting in China. From a geographical perspective, AI patenting efforts remain highly concentrated, with inventors in the top 10 countries contributing almost 96% of all worldwide AI patent applications.
The USA, China and Japan are among the most patent-intensive and specialized countries in the AI domain, followed by South Korea, Germany, and the UK. Over the past three decades, the USA has maintained a leading position, with China growing most rapidly in recent years and Japan seeing relative loss of its early leadership position.
Most AI patents are owned by large private companies. Public organizations and universities are less prominent in the ownership of AI patent applications, except in China where public research organization and universities are frequent AI patent assignees.
AI patents cover a wide range of technological areas, not only in information technologies and computing but also in applied fields such as medicine, healthcare, finance, and education. The authors’ analysis finds that flows of technological knowledge tend to be concentrated among patents developed in the same country, mostly dominated by the USA although there were some technological routes led by China.
The paper, “Mapping technological innovation dynamics in artificial intelligence domains: Evidence from a global patent analysis” by Na Liu, Philip Shapira, Xiaoxu Yue, and Jiancheng Guan (PLoS ONE 16(12): e0262050) is open access and downloadable at https://doi.org/10.1371/journal.pone.0262050.