인공지능 관련 기술의 특허 보호 전략–미국 특허법을 중심으로

The advancement of artificial intelligence technology is reshaping all industries by mimicking and many times surpassing human intelligence. As artificial intelligence begins to play a central role as a creator and contributor to creativity and innovation, it raises new challenges for the current patent system, which is centered around human-only creative acts. This study focuses on current issues in U.S. patent law related to artificial intelligence technology, including subject matter eligibility, written description and enablement, non-obviousness, and issues related to datasets. In the context of the strict patent subject matter eligibility test in the U.S., this study aims to derive strategies to navigate artificial intelligence technology that mimics human’s mental process. In addressing written description and enablement requirements, the scope of disclosure can vary by the level of intelligence performed by the artificial intelligence system seeking patent protection. The artificial intelligent systems that are capable of performing high-dimensional predictive functions without human guidance may require more disclosure of algorithms and data to receive patent protection. Concerning the non-obviousness requirement, we will explore R&D environments where human researchers and articial intellegence collaborate throughout the inventing process. “A person having an ordinary skill in the art” in the non-obviousness requirement is no longer a person of ordinary skills, but the collaboration between a person and artificial intelligence system, makes a person with extraordinary skills with capacity of artificial intelligence. Then, all inventions from R&D processes will become predictable and obvious to the extra ordinary person having the capacity of artificial intelligence. This study also delves into the protection of data, a crucial ingredient in artificial intelligence technology, examining strategies to obtain patent protection and exploring trade secret as an alternative. Ultimately, this study aims to derive patent protection strategies for artificial intelligence related inventions in the U.S. and Korea by analyzing U.S. case laws and patent office examination guidelines of both countries.