Patent Eligibility of AI-Related Inventions in India

Introduction

Since 2013, almost 170,000 AI-related patents have been filed globally, accounting for half of all AI-related patents ever filed¹.

Unlike other major jurisdictions like Europe, Japan, and the US, India has not issued specific rules for examining inventions related to artificial intelligence. The Computer-Related Inventions Guidelines 2017 (CRI guidelines)² are used to evaluate these inventions. In other words, the subject matter exclusions stated in Section 3(k) of the Indian Patents Act, 1970 (hereinafter “the Act”) are applied to assess inventions including artificial intelligence³.

This article discusses how AI-related inventions are examined for subject matter eligibility in India.

AI Patent Filing Trends in India

Table 1 represents the descriptions of the most frequently used IPC codes in AI-related patents⁴:

Table 1

IPC Codes	Description
G06K	Recognition of data, presentation of data, record carriers; handling record carriers
G06K 9/00	Methods or arrangements for reading or recognising printed or written characters or for recognising patterns
G06K 9/62	Physics > computing; calculating; counting > recognition of data; presentation of data; record carriers; handling record carriers > methods or arrangements for reading or recognizing printed or written characters or for recognizing patterns, e.g. Fingerprints > methods or arrangements for recognition using electronic means
G06T	Image data processing or generation
G06T 7/00	Image analysis
G06T 19/00	Manipulating 3D models or images for computer graphics
G06F	Electric digital data processing
G06F 3/01	Input arrangements or combined input and output arrangements for interaction between user and computer
G06F 7/00	Physics > computing; calculating; counting > electric digital data processing > methods or arrangements for processing data by operating upon the order or content of the data handled
G06F 17/00	digital computing or data processing equipment or methods, specially adapted for specific functions
G06F 17/30	Information retrieval; Database structures therefor
G06F 19/00	Physics > computing; calculating; counting > electric digital data processing > digital computing or data processing equipment or methods, specially adapted for specific applications
G06N	Computer systems based on specific computational models
G06N 3/04	Architecture, e.g. interconnection topology
G06N 3/08	Learning methods
G06N 20/00	Machine learning
G10L	Speech analysis of synthesis; speech or voice processing; speech or audio coding or decoding
G10L 15/22	Procedures used during a speech recognition process, e.g. man-machine dialog
H04N	Pictorial communication; television
H04N 5/232	Devices for controlling television cameras
H04L	Electricity> electric communication technique> transmission of digital information, e.g. Telegraphic communication
H04L 29/06	Electricity> electric communication technique> transmission of digital information, e.g. Telegraphic communication > arrangements, apparatus, circuits or systems, not covered by a single one of groups > characterized by a protocol
H04L 29/08	Electricity > electric communication technique > transmission of digital information, e.g. Telegraphic communication > arrangements, apparatus, circuits or systems, not covered by a single one of groups > transmission control procedure, e.g. Data link level control procedure
A61B	Human necessities > medical or veterinary science; hygiene > diagnosis; surgery; identification
A61B 5/00	Human necessities > medical or veterinary science; hygiene > diagnosis; surgery; identification > Measuring for diagnostic purposes; identification of persons

According to a new report by the World Intellectual Property Organization (WIPO), India ranks fifth overall in terms of the number of Generative AI (GenAI) inventions patented, behind China (38,210), the US (6,276), the Republic of Korea (4,155), and Japan (3,409). In terms of the total number of published GenAI patents, India has surpassed the UK (sixth place with 714 patents), Germany (708), and other countries. Having said that, the data shows that India has filed only 1,350 patents for GenAI inventions between 2014 and 2023.

At the same time, according to the Report, India has the fastest yearly growth rate in GenAI patent publications, with an average rate of 56%, indicating the country’s fast-paced innovation environment. This is based on a small base, though, as India has only 3% of all GenAI patents⁵.

Eligibility Criteria for Patentability of AI Inventions

For an AI-related invention to be eligible for patent protection, it must satisfy the essential legal conditions of novelty, inventive step, and industrial applicability, just like any other invention. The Indian Patents Act does not provide a definitive definition of what constitutes an invention. Instead, it offers a non-exhaustive list of “non-inventions” under Section 3, relevant to us are⁶:

Section 3(k): “A mathematical or business method or a computer programme per se or algorithms” is not patentable; and

Section 3(m): “A mere scheme or rule or method of performing mental acts or a method of playing a game” is not patentable.

For computer-related inventions, there is a detailed guideline called CRI Guideline which states it is only when contribution to an invention solely lies in the computer programme per se or algorithm or business method or mathematical method, the claimed invention cannot be allowed for falling under Section 3(k) of the Act. While establishing patentability, the focus must be on the underlying substance of the invention and not on the particular form in which it is claimed. If in substance, the claim, taken as a whole, does not fall in any of the excluded categories, the patent should not be denied.

The requirement to establish the “technical contribution” and “technical effect” of the claimed subject matter has also been established by various case laws including Ferid Allani⁷ (“Innovation in the field of artificial intelligence, blockchain technologies and other digital products would be based on computer programs, however the same would not become non- patentable inventions – simply for that reason. It is rare to see a product which is not based on a computer program. Whether they are cars and other automobiles, microwave ovens, washing machines, refrigerators, they all have some sort of computer programs in-built in them. Thus, the effect that such programs produce including in digital and electronic products is crucial in determining the test of patentability”). and Microsoft⁸ (“If a computer-based invention provides a technical effect or contribution, it may still be patentable. The technical effect or contribution can be demonstrated by showing that the invention solves a technical problem, enhances a technical process, or has some other technical benefit. The mere fact that an invention involves a mathematical or computer-based method does not automatically exclude it from being patentable. The invention can still satisfy the patentability requirements, including the requirement for a technical effect or contribution, to be eligible for patent protection.”)

Thus, the bar on patenting is in respect of ‘computer programs per se….’ and not all inventions based on computer programs. The technical effect that invention produces is crucial in determining the test of patentability. Sometimes the computer program may include certain other things, ancillary thereto or developed thereon, which may be eligible. Computer programmers as such are not eligible. The words “per se” were incorporated in Section 3(k) to ensure that genuine inventions which are developed, based on computer programs are not refused patents.

Sample Granted AI-Related Patents

Here are some examples of the patents that have received grants in India with a specification of demonstrating technical effects:

Indian Patent Application No. 60/CHE/2010, titled “Artificial Intelligence-Based Assessment and Tutoring System and Method Thereof”, was granted on demonstration of technical solution viz. contextual assistance in the form of audio and video, using dynamic speech processing with emotional elements such as encouragement tone to enforce learning.

Indian Patent Application No. 202041003260, titled “Artificial Intelligence Enabled Automated System for Targeted Content on Display Devices”, was granted upon a demonstration of the technical advancement in solving the issues related to the conventional system by using classification, scene understanding using deep learning architectures such as convolutional neural network and blockchain module which is used for recording information in an unalterable manner. The impugned invention achieved the objective of displaying the targeted content to a targeted audience on real real-time basis.

Indian Patent Application No.: 202014026775, titled “Artificial Intelligence Processor and Method of Performing Neural Network Operation Thereof”, was granted upon a demonstration of the technical effect of reducing resource or power consumption of hardware of a terminal device required for image processing.

Given the above, it is clear that to make AI-related inventions qualify as a patentable subject matter, and avoid its fall under excluded subject matter, it is required to significantly integrate essential elements or features of AI in a practical aspect. If any invention about AI shows technical advancement and is carried out by structural components, it would be considered patentable subject matter.

Our Recommendations for Patenting AI-related Inventions⁹

If the patent is for a hardware-based innovation, each structural component of the invention must be elucidated in detail with appropriate illustrations. The operating relationship between various components must be defined.
If the invention relates to a method, the necessary sequence of steps must be clearly stated to separate the invention from the previous art, using flowcharts and other information essential to carry out the invention, as well as their modes/means of implementation.
The physical and constructional features which are vital to the functioning of the invention should be mentioned in the claims.
The operation of an AI-powered device should be clearly described.
Claiming AI programming codes/ algorithms, computer-readable storage medium/ computer programs should be avoided.
Means plus function claims should be avoided.

Conclusion

Till now there are no specific guidelines for AI-related inventions in India, unlike other Jurisdictions like the US, EP, and Japan which have specific guidelines. Until specific guidelines for AI-related inventions are issued in India, according to the current CRI guidelines and court precedents, technical advancement and structural features are essential for an AI-related invention to be granted patent protection in India.