국내외 인공지능의 특허 동향 분석

Artificial intelligence (AI) technology, a key driver of the 4th industrial revolution, is emerging as a central axis of the global technology competition, and the technology gap between countries is acting as an important factor in determining future economic dominance. This study comprehensively analyzed each country's technology competitiveness and development strategies for 12 years from 2012 to October 2023 through domestic and international patent application trends in this AI technology competition. As a research methodology, we established a three-step systematic search strategy using CPC (Collaborative Pattern Classification) code and 600 detailed keywords for IP5 countries (Korea, the United States, Europe, Japan, and China), and selected and analyzed the final effective patent through verification and deduplication of the relevance of the first collected patent technology. As a result of the study, China showed an overwhelming share of 40.2% of all AI-related patent applications, recording an annual growth rate of 48.3%, followed by the U.S. 28.3%, Europe 12.1%, Korea 10.7%, and Japan 8.7%. By technology field, learning and reasoning ability accounted for the highest proportion with 36.8%, followed by visual intelligence (28.9%), and auditory intelligence (15.2%). In particular, since the introduction of transformer technology in 2017, the natural language processing sector has grown 67.3% annually, and since the launch of ChatGPT in 2022, the technology innovation cycle and patent application trends have increased 156% year-on-year, showing a close correlation. Although Korea showed relatively strong strengths in the field of visual and auditory intelligence, such as establishing a patent portfolio centered on large companies such as Samsung Electronics and LG Electronics, the proportion of overseas patent applications was 28.1%, which is significantly lower than that of the U.S. (51.8%), Japan (63.8%), and Europe (85.3%), indicating that a strategic approach to preoccupy the global market is urgently needed. In addition, Korea's patent quality index is 8.4, which is relatively lower than that of the U.S. (9.5 points) and Europe (9.1 points), indicating that quantitative growth and qualitative improvement are needed. Policy implications include first, intensive investment to bridge the technology gap between Generative AI and large language models, second, strengthening global IP strategies by expanding the support system for overseas patent applications, and third, promoting commercialization of basic research results through industry-academia cooperation. This study objectively evaluated each country's technological competitiveness through quantitative analysis of AI patent data and suggested future policy directions, and confirmed that patent data alone has limitations in that it is difficult to fully grasp the actual technology implementation or commercialization level.