ソフトバンクグループの創業者・孫正義氏は、ChatGPTとの壁打ちでわずかな時間で数百もの先進的アイデアを創出・発展させ,1000件以上の特許出願を行っているということです。
生成AIの急速な進化は、特許審査における進歩性(非容易想到性)の判断にもさまざまな影響を及ぼす可能性があります。以下では、その主な影響をいくつかの観点から整理します。 1. 先行技術調査・引用可能な情報の増大
2. 当業者の技術水準の向上と「容易想到性」の基準変化
3. 知的財産庁の審査プロセスへのAIの活用
4. 特許性のハードルの高まりと新たな戦略
まとめ 生成AIの急速な進化により、特許審査における進歩性の判断基準は今後厳格化していく可能性があります。AIが先行技術の調査やアイデア生成を容易にするほど、当業者にとって「自明かどうか」の線引きがシビアになる一方、特許庁側もAIを活用して審査プロセスを高度化していくと考えられます。 その結果、発明者や企業は「AIでは簡単に到達できないレベルの技術的特徴」をどのように示すかがより重要となり、明確かつ説得力のある特許明細書の作成や、ノウハウの秘匿化など、多角的な知財戦略が求められるようになるでしょう。 (本稿は、生成AIと共に書き起こしたものです。) The Impact of Generative AI on the Standards for Assessing Inventive Step in Patent Examination Introduction Masayoshi Son, founder of the SoftBank Group, reportedly used ChatGPT to brainstorm hundreds of advanced ideas in a short period, leading to over 1,000 patent applications. The rapid evolution of generative AI could significantly influence how inventive step (non-obviousness) is assessed during patent examinations. Below, we outline the main impacts from several perspectives. 1. Increased Prior Art and Referable Information • Explosive Growth in AI-Generated Ideas Advancements in generative AI enable the automatic generation and publication of diverse ideas and technical combinations at unprecedented speed. This may result in a substantial increase in “prior art” available for citation by patent offices. As a consequence, patent offices are expected to conduct broader and more rapid prior art searches, potentially leading to a stricter evaluation of whether a claimed invention is “obvious to a person skilled in the art.” • Challenges from Information Overload Conversely, the abundance of information may increase the risk of examiners overlooking critical prior art or facing difficulties in conducting accurate evaluations. To address such challenges, patent offices are likely to adopt AI tools for efficient and comprehensive searches and analyses. 2. Higher Technical Standards for Practitioners and Changes in the Definition of “Non-Obviousness” • Proliferation of Advanced Tools The widespread adoption of generative AI could raise the perceived skill level and knowledge base of the “person skilled in the art.” Consequently, the threshold for determining “non-obviousness” under patent law may become more stringent. • Treatment of Combination Inventions Generative AI excels at combining existing technical ideas to propose novel solutions. While combination inventions have historically been scrutinized based on whether the combination is obvious, AI’s ability to create “clever combinations” could make it increasingly challenging for examiners to distinguish between obviousness and inventive advancement. 3. Use of AI in Patent Examination Processes • AI as a Support Tool for Examiners Patent offices are accelerating efforts to integrate generative AI and natural language processing technologies. These tools enable:
In May 2024, the "Interim Report of the Study Group on Intellectual Property Rights in the AI Era," compiled by the Study Group on Intellectual Property Rights in the AI Era, stated: "At present, there are no circumstances that warrant changes to the operation of patent examination practices due to the impact of AI utilization in the invention and creation process. Therefore, when assessing inventive step, it is considered appropriate to set and determine the level of inventiveness based on the established practice, while accurately understanding common technical knowledge and technical standards, including the utilization of AI in the invention and creation processes across a wide range of technical fields." As such, a review of the examination standards is unlikely in the immediate future. However, as AI utilization advances, continuous reassessment of how the "ordinary creative ability of a person skilled in the art" is positioned will likely take place. In rapidly evolving technological fields, it is also expected that the guidelines and examination standards provided by the Patent Office will be periodically revised. 4. Higher Thresholds for Patentability and Emerging Strategies • Importance of Truly Innovative Elements In an era where AI generates an abundance of ideas, identifying “genuinely innovative elements” that cannot be easily conceived using AI will become key to securing patents. To argue for inventive step, applicants must clearly articulate technical features and outcomes that exceed AI’s generative capabilities. • Changes in Intellectual Property Strategies The spread of generative AI could lead to previously patentable inventions being deemed “obvious.” In response, applicants may need to adopt new approaches, such as:
Conclusion The rapid evolution of generative AI may lead to stricter standards for assessing inventive step in patent examinations. As AI facilitates prior art searches and idea generation, the distinction between what is “obvious” will become increasingly nuanced. Meanwhile, patent offices are expected to leverage AI to enhance examination processes. For inventors and companies, demonstrating technical features beyond AI’s reach will become crucial. This will necessitate crafting clear and persuasive patent specifications and adopting multifaceted intellectual property strategies, such as maintaining trade secrets. (This article was drafted with the assistance of generative AI.)
0 Comments
Leave a Reply. |
著者萬秀憲 アーカイブ
January 2025
カテゴリー |