The Japan Patent Office has recently published useful materials on AI-related inventions. Taken together, these materials provide a helpful picture of two important developments in Japan. One is the continuing growth of AI-related patent filings. The other is the JPO’s effort to clarify examination practice through published case examples. For foreign applicants and overseas patent professionals, these publications offer a valuable overview of both filing trends and practical examination issues in Japan.
Recent trends in AI-related inventions
According to the JPO’s recent survey, patent applications relating to AI-related inventions in Japan have increased significantly since 2014, reaching approximately 11,400 in 2023. While so-called “AI Core Inventions” remain important, the broader trend suggests that AI technologies are now being applied across a wide range of technical and business fields. The survey points to active filing trends in such areas as AI core technologies, image processing, image recognition, healthcare, information retrieval, and natural language processing.
The survey also notes a notable increase in business-related inventions, particularly in classifications associated with digital services and data-based business methods. This appears to reflect the wider commercial use of AI, including recent developments in generative AI. Another interesting point is the change in technical focus: applications referring to convolutional neural networks (CNNs), recurrent neural networks (RNNs), and long short-term memory networks (LSTMs) increased until around 2021 and then began to decline, whereas applications referring to transformer models have continued to increase. These developments suggest that Japan is seeing not only greater filing activity in AI, but also a shift in the underlying technologies being claimed.
For further details, please see the JPO’s publication on recent trends in AI-related inventions.
Patent examination case examples for AI-related inventions
The JPO has also published examination case examples concerning AI-related technologies. Rather than creating a special set of patent rules only for AI, the JPO explains how existing principles of Japanese patent law apply to AI-related inventions. This approach is especially helpful for applicants and practitioners seeking a clearer understanding of how AI-related claims may be examined in practice.
The published materials include case examples addressing three main issues: description requirements, inventive step, and patent eligibility. The latest compilation contains 25 case examples in total, including more recent examples involving generative AI and large language model-related technologies. These examples show how the JPO analyzes issues such as whether the claimed invention is supported by the specification, whether a claimed feature represents a technical contribution rather than a mere abstract idea or business arrangement, and how inventive step may be assessed in AI-related contexts.
From a practical perspective, these materials are useful not only for prosecution but also for drafting. They suggest the importance of clearly explaining the role of data, the technical function of AI processing, and the relationship between the claimed invention and the problem to be solved. For foreign applicants considering patent protection in Japan, the case examples provide helpful guidance on how AI-related applications may be better prepared for Japanese examination practice.
For further details, please see the JPO’s publication on patent examination case examples pertinent to AI-related technologies.
Concluding remarks
These recent JPO publications are worth reading together. One helps readers understand how AI-related patent activity is developing in Japan, while the other provides a practical view of how AI-related inventions are examined. For overseas applicants, this combination offers useful guidance both for filing strategy and for the preparation of stronger patent applications in Japan.
