Demystifying Rights Of AI Generated Inventions

AI is increasingly present in all areas of the economy and our daily lives, for example, an open AI chat bot – ChatGPT or Google search engine. The broad use of AI can only be secure if it is founded on suitable legal regulation; capable of understanding and limiting its impact on daily life; and ensures suitable enforcement.

Artificial Intelligence as an Inventor of Patents

To start with, it is first worth defining what AI is. AI is defined by two basic concepts: autonomy and adaptivity. Autonomy in AI means that the AI ​​can perform tasks in complex environments without constantly following user instructions, whilst adaptivity in this context means that AI has the ability to improve performance by learning from experience.

In general, artificial intelligence (AI) refers to systems that display intelligent behaviour by analysing their environment and automatically taking some degree of actions to achieve specific goals.

Basic requirements to qualify as an Inventor in India:

One of the biggest challenges of AI innovation is that it has the potential to completely change the nature of innovation. In this respect, there are several important aspects to consider while dealing with the issues related to inventorship of AI.

Section 2 and Section 6 of the Indian Patent Act, 1970 elaborates the criteria for recognizing an inventor and the person as an applicant who can file for a patent in the Indian jurisdiction. In particular, Section 6 provides information regarding a person entitled to apply for a patent. Section 2(1)(s) provides information regarding who can be referred to as a ‘Person', whilst Section 2(1)(y) discloses those who cannot be a true and first inventor while filing a patent application.

Additionally, it is pertinent to note that Section 2(1)(s) of the Act clearly states that the person filing for a patent may be either a ‘natural person’; or any ‘Government’ Organization. However, it is interesting to note that the term ‘person' is not merely limited to any natural person and/or Government but also includes the reference to Government along with other entities that can claim to be the true and first inventor of the invention. It can also be construed that anybody or anything that is the first and true inventor of the invention qualifies as the ‘person' for whom a patent application is being submitted.

Although Section 2(1)(y) mentions who cannot be an inventor, is unclear who can be considered the actual and original inventor of the AI-generated invention. Thus, the below mentioned Indian case law with combination of the Patent Act specifies the basic requirement for being an inventor in India.

Further, the Ayyangar Committee Report of 1959 discloses the legislative intent behind the Patent Act, where the report addresses the intent of mentioning an inventor as a matter of right of the inventor for a particular patent. The report emphasizes that any person who has a moral right to be named as an inventor, even though he might not have all the legal rights over the invention. According to the report, the idea of approaching an inventor about their creation is to help the inventors increase the financial value to which they are legally entitled, even if they may have given up their exclusive rights through contracts or agreements. However, the AI can neither luxuriate in the benefits envisioned in the legislative intent, nor it can entitle the moral rights under prevailing Indian laws or enforce misuse.

Although the current international, Indian and various national regulations clearly state that the inventor must be a natural person. In this regard, it is worth considering on the extent to which this provision will serve the primary purpose of the patent system in the social and economic conditions of the near future. The purpose of the patent system is to make inventions widely known and useful to society through compulsory publication, and to ensure that the knowledge becomes known to everyone and useful to society, so that the technology can further be developed. The right holder acquires an exclusive right in the market for a limited period under certain conditions for such patented invention, and thus an economic interest attached to it is also created. Therefore, it is necessary to think what if in the future, inventions created by AI are removed from the scope of protection. However, there is a high risk that a lot of knowledge can remain a trade secret for AI-generated invention, which could have the effect of slowing down the development of technology.

Ascertaining the “Inventor” in AI invention:

Patent Application naming AI as an Inventor - DABUS

More than a dozen nations, including the European Union, received two patent applications in 2019 with DABUS (Device for the Autonomous Bootstrapping of Unified Sentience) listed as the inventor. The applications listing DABUS as the inventor was claimed in the name of a natural person as an applicant – Dr. Stephen Thaler. It purportedly came up with two distinct inventions without any help from humans, and as a result, it was listed as an inventor on patent applications for both inventions. The idea of assigning inventorship to an AI-machine not only created new legal challenges, but also raised global intellectual property (IP) concerns over whether an AI machine can be named as the inventor of its AI-related patents.

The South African Patent Office and the Federal Court of Australia are the two jurisdictions where the DABUS application was filed that have so far recognized and accepted DABUS as an inventor, whereas the European Patent Office (EPO), the United Kingdom Intellectual Property Office (UKIPO), the United States Patent & Trademark Office (USPTO), and other IP offices have not, by asserting that an inventor must be a natural person.

The following segments gives a brief overview of the AI-inventorship related to the decision in various IP offices.

South Africa: The South African Patent Office became the world's first IP office to grant patent for an invention developed by the AI ​​machine DABUS. However, it is pertinent to note that South African patent law does not define "inventor".

Australia: On July 2021, the Australian Federal Court overturned the Australian Patent Office's refusal to grant inventorship to AI machines, ruling that artificial intelligence systems could be an inventors. This decision provided the following rationale:

• an agent can be an inventing person or thing, where an agent can be described as computer, regulator, dishwasher and like; and
• no specific provision in the Australian Patent Act prevents an AI system from being considered as an inventor.

The court in its judgement cleared that the ruling affects only to the inventorship of patents and not the ownership.

Europe and United Kingdom: Unlike South Africa and Australia, the European Patent Office (EPO) and United Kingdom Intellectual Patent Office (UKIPO) have refused to grant inventorship rights to an AI machine. The EPO reasoned that the term "inventor" in European Patent Convention refers only to a natural person. The EPO also stated that merely giving a machine a name does not endow the machine with a legal personality as some legal statuses associated with the title of "inventor" require a legal personality to exercise. The UKIPO made its judgment based on justifications echoed by the EPO.

United States: The United States Patent and Trademark Office (USPTO) also rejected DABUS reasoning that a plain reading of statutory provisions of United States Code (U.S.C.) discloses the inventors only as natural persons. The USPTO referred to 35 U.S.C. Section 100(f)-(g) and Section 101, where terms such as “individual” and “whoever” are used in the context of inventorship, stating that inventors have to be natural persons. In accordance with the 35 U.S.C. Section 115(b), a person who believes "himself or herself" to be the inventor of the claimed invention must sign an oath or declaration. Additionally, the USPTO cited a number of Federal Circuit decisions that said only natural persons could be the inventors. For example, In University of Utah v. Max Planck Gesellschaft zur Forderung der Wissenschaften E.V, 734 F.3d 1315, 1323 (Fed. Cir. 2013), the Federal Circuit refused to list a company or state as an inventor as the USPTO concluded that an AI system cannot be listed as an inventor and also the relevant statutes limit inventors to individuals.

Japan: According to the Japan Patent Office (JPO), AI policies that use machine learning or deep learning to achieve better results with slightly modified algorithms are just viewed as a routine upgrade unless it shows that this method was never applied before. Therefore, to make the patent application acceptable in Japan, the AI ​​invention must show that the method was not known in any form of prior art, and that it is not just a mere improvement over any prior art.

India : In India, the Controller General of Patents recently recorded objections against AI-generated invention (DABUS) having patent application numbered 202017019068. The examination report cited objection to the DABUS patent application under Section 2 and Section 6 of the Patents Act 1970 (“Act”). The Controller General of Patents raised an objection in the Examination Report, stating that application lacks from passing formal and technical review under Sections 2 and 6 of the Patents Act, 1970 – as DABUS is not recognized as a person.

The Controller General of Patents laid out objections in the Examination Report of Thaler's Indian patent application, stating that the application could not pass formal and technical examination under Section 2 and Section 6 of the Patents Act, 1970 – as DABUS is not recognized as a person. The same was supported by a number of legal precedents, such as in V.B. Mohammed Ibrahim v. Alfred Schafranek, where the Court ruled that neither a corporation nor a financing partner can be the sole inventor, and inter alia held that only a natural person who actually contributes their skill and knowledge to the innovation is able to legally claim the inventorship.

The applicant responded to the aforementioned objections raised in FER regarding the inventorship of AI by asserting that DABUS is the true inventor/devisor of the invention and that, in accordance with the Indian Patent Act, the applicant have named the inventor/devisor of the invention at the time the application was filed. The definition of inventorship in various jurisdictional procedures is geared toward natural beings with the intention of preventing company invention, according to the response's further explanation of the notion. It was not the outcome of a careful consideration of autonomous machine creation, hence it should not be prohibited from retaining intellectual property rights in cases where no normal person meets the requirements of an inventor.

On the other hand, it may be claimed that an AI may also contribute its technical skills to an invention in order for it to qualify as an inventor. Another Indian case law can support the same claim: Som Prakash Rekhi v. Union of India & Anr on November 13, 1980 AIR 1981 SC 212, the Hon'ble Supreme Court of India rendered a decision on what constitutes a "person" in the eyes of the law. The judgment concluded as a result that a jurisdictional person is the one to whom the Law considers ‘personality'. When we refer to a juristic personality, we actually refer to a legal entity who have rights to sue or who can be sued by another entity. Inherently, an AI does not have the capability to use the numerous rights, nor can it perform the required duties of any legal entity independently.

Impact of AI Tools like ChatGPT on Patent Applications

A key event that has sparked a lot of the current debate was with the launch of on open AI-system, called ChatGPT (Chat Generative Pre-trained Transformer) in November 2022. ChatGPT is an AI chatbot developed to mimic human conversations through the use of a large language model. ChatGPT is able to answer exam questions; write songs, short stories and poems; and even generate and debug computer code, all from relatively simple text inputs.

However, there is widespread concern that these text-generation tools could be able to replace or handle the complications of patent law. In regard to this, it is also been tried and confirmed by various patent professionals that ChatGPT lacks the legal knowledge necessary to render legal advice or draft a patent application that would be legally binding. Even in some of the technologies related to non-patentable subject matter, ChatGPT failed to suggest the same. Although, AI technologies like ChatGPT can be very helpful in spotting errors in applications, including antecedent basis and improperly used reference numbers, as well as in enhancing the quality of patent applications. The AI language model of ChatGPT can be of valuable support in refining the writing, and catching grammatical and syntax errors, but we still cannot be solely dependent on it for drafting of any application.

Hence, the speed and precision of the job can be significantly impacted by the incorporation of AI technologies like ChatGPT into the patent drafting process. Yet it's crucial to keep in mind that ChatGPT is not a replacement for a registered patent attorney or agent. It is a tool to be utilized in conjunction with the expertise and knowledge of a patent professional. Combining the strength of AI with legal expertise of a patent professional delivers the best of both worlds, streamlining the patent writing process, reducing time spent, and ultimately delivering higher-quality patent applications.

Role of Artificial Intelligence in Economic Growth in India

The Parliamentary Standing Committee (“Committee”) constituted under the Department of Commerce, analysed the current landscape of the IPR regime in India and noted its contribution to fostering innovation and entrepreneurship in the country in its report titled as “Report 161: Review of the Intellectual Property Rights Regime in India” presented in the Rajya Sabha on July 23^rd, 2021. Particularly, the report examined the existing challenges related to the current legislative structure including the inventorship rights of an AI. The same report was also taken as a support in the response to the objections of first examination report of Indian application, filed by Dr. Stephen Thaler.

The committee recognized the relevance and usefulness of AI-based cutting-edge technologies and machine learning. Digital technology is proving essential in responding to the global crisis, especially in this time of great impact from the pandemic. Further, the Committee placed reliance on a report published by Accenture titled as “How AI Boosts Industry Profits and Innovation” which estimated AI to add US $ 957 Billion into the Indian Economy by 2035, if used optimally, give a better understanding of the impact and role of AI and technology increase in relation to the contemporary landscape and intellectual property.

The Commission therefore recommended giving priority consideration to the relevant provisions of the Indian Patents Act, 1970 [i.e. Section 3(k)], and the Copyright Act, 1957 to grant invention rights to AI in India. The report also stated that “The Committee recommends the Department that the approach in linking the mathematical methods or algorithms to a tangible technical device or a practical application should be adopted in India for facilitating their patents as being done in the EU and U.S. Hence, the conversion of mathematical methods and algorithms to a process in this way would make it easier to protect them as patents“.

The Committee came to the conclusion that the changes to the legal framework would safeguard AI creations (whether created autonomously or with human assistance or input), encourage ground-breaking research and development in the nation, and keep the environment favorable for the protection of innovations involving human intelligence. The Committee argued that because such AI-induced innovations would not be protected domestically, the restrictions placed on the inventorship rights of an AI would deter large investments in the industry.

Threat to Human-Devised Invention:

Automatic language translation services, chatbots for customer support, voice and facial recognition systems, and virtual assistants are just a few examples of the many areas of society where data-driven AI is quickly becoming established. But much like every great discovery, improvements in AI also raise significant questions and uncertainty about societal and economic structures.

AI is currently actually putting the core of innovation at jeopardy. Research is the usual starting point for an invention, which leads to patents and other benefits. Yet, AI appears to be able to complete every task on its own, which can reduce the incentives for advances and raise concerns about the recipients. The existential issues with ownership, inventorship, and infringement, among other fundamental principles of the patent system, were undoubtedly discovered by AI, and the DABUS case aggravated such an issue.

The Way Forward for Inventorship Rights of an AI System

The particular terminology in each jurisdiction's patent laws will determine whether an AI qualifies as an inventor under patent law. In countries where clear statutory text requires an inventor to be a normal person, legislative modifications and amendments may be necessary to expressly incorporate and recognize AI-generated ideas. In countries like Australia where the statutory language is less clear, the courts may be given more latitude to take deliberate statutory interpretation and policy factors into account. Therefore, a common strategy is expected to be adopted by all IP offices to successfully meet the emerging difficulties posed by inventions by AI.

Authors: Renu Bala Rampal (Associate Partner) and Swaraj Singh Raghuwanshi (Senior Associate) at S.S. Rana & Co. Views are Personal.

For more information, please write to info@ssrana.com