The China National Intellectual Property Administration ("CNIPA") issued the Guidelines for Patent Applications for Artificial Intelligence ("AI") Related Inventions (Trial Implementation) ("Guidelines") on 31 December 2024, aiming to clarify and specify China's patent examination standards in the AI field and address the key legal issues of concern to AI innovators.
HIGHLIGHTS
1. Four Categories of AI-Related Patent Applications
Based on the "role" that AI plays in the invention, AI-related patent applications are classified into 4 categories:
AI algorithms or models themselves;
functions or field applications based on AI algorithms or models;
inventions made with AI assistance (with substantial human contribution); and
AI-generated inventions (without substantial human contribution, and presumably not eligible for patent protection – see below "Inventorship").
2. Five Key Legal Issues
Inventorship: For inventions made with AI assistance and AI-generated inventions, the Guidelines require that inventors must be natural persons and not AI systems or other non-natural persons. This is in line with the CNIPA's current Patent Examination Guidelines which require inventors to be individuals and not entities or AI. This is so even if there are multiple inventors. The rationale is that only "civil subjects" under the civil law can enjoy the civil rights of inventors and AI systems are currently not capable of being civil subjects and enjoying civil rights. This position is also in line with the current approaches by most of the other major jurisdictions including the US, the UK, the European Patent Office, Japan, Australia, etc., e.g. in their judgments/decisions related to the DABUS patent applications in recent years.
Patentable Subject Matters: To be eligible for patent protection, an AI-related solution (i) cannot merely involve rules and methods of intellectual activities (Article 25.1(2) of the Patent Law) (e.g. merely involving abstract mathematical theories or algorithms without any technical features), and (ii) must be a technical solution (Article 2.2 of the Patent Law) which reflects the adoption of technical means that follow the laws of nature to solve a technical problem, and which achieves a technical effect (e.g. improving the internal performance of the computer system, etc.) that conforms with the laws of nature.
Sufficient Disclosure of Invention in Patent Description: This requirement is particularly important for AI-related patent applications due to the "black box" nature of AI algorithms or models, and the Guidelines require clear and complete disclosure of the necessary technical content of the AI-related invention so as to enable a person skilled in the art to implement the invention. Such necessary technical content varies depending on different inventions, but the patent description should sufficiently describe the part which made contribution to the prior art. The Guidelines also provide some drafting suggestions for different types of AI-related inventions.
Inventiveness: To determine whether an AI-related invention possesses sufficient "inventiveness" to qualify for patent protection, holistic consideration should be given to both technical features and algorithmic features that have an inter-supportive and interactive relationship with the technical features. A technical solution has "inventiveness", if, upon holistic consideration, it has prominent substantive features and represents a notable progress as compared to the prior art.
Ethical Issues: Last but not least, the Guidelines briefly discuss the ethical issues associated with AI-related patent applications – namely, reiterating the necessary compliance with laws, social ethics and public interests under Article 5 of the Patent Law, particularly around data acquisition and utilization by AI where care should be placed on the source, application scenarios, security management and usage of the data, as well as data content and the specific methods of data collection, storage and processing.
TAKEAWAYS
The Guidelines provide important practical guidance to AI businesses, patent law practitioners and other stakeholders under China's current patent law framework. On one hand, the Guidelines demonstrate China's commitment to protecting (certain types of) AI-related inventions and fostering AI-related innovations. On the other hand, by clarifying that AI systems are not yet capable of being inventors, the Guidelines attempt to strike a balance between technological innovation and existing legal norms.
As AI rapidly develops, the interaction between AI and intellectual property is also evolving. There will undoubtedly be further developments in this space as the Chinese authorities build a framework which enables the development of AI in China within the context of an existing IP system which emphasizes, amongst other things, traditional concepts of inventorship by natural persons and sufficient disclosure of the invention. We will continue to monitor updates in this space.
[View source.]
Comentários