Introduction

Artificial intelligence is generally known as AI is the latest buzz around the corner, not very long ago it was considered fiction and helped many movie creators to find a good movie plot where the machines were shown to have the brain power and would one day make the entire human race as their slaves. However, it is a reality now and perhaps almost every faculty in the field of science and technology use these software modules to generate sophisticated work. In 2020, an ML (machine learning) algorithm helped researchers to develop a powerful antibiotic that works against a broad spectrum of pathogens.1 AI and ML are being extensively used to develop new vaccines, drugs, materials discovery, space technology, automobile technology, industrial automation and many more. This paradigm shift has fueled a new debate concerning the severe impact on intellectual property laws.

As technology evolves, unnatural intelligence (AI / ML) created discoveries that are, innovations brought either automatically or semi-automatically by AI and ML modules are regarded to be becoming more widespread. The human imagination in such discoveries is less to nothing, while at the same time the act of inventing turns out to be a lot simpler, it is because most of the mental exercise involved is performed by the automated machines via ML and AI.

Just recently something astonishing has occurred in the world of IPR, the patent office in South Africa has for the first time ever in the history of intellectual property granted a patent to an AI. Yes, you read that right, a Patent where the creator is an Artificial Intelligence.

"DABUS" (Device for Autonomous Bootstrapping of Unified Sentience) is the AI module that has made this happen. While there is an agreement that all such inventions from AI and ML modules are nearly contradictory to the concept of human inventorship, it remains largely unclear to what extent concerns regarding 'non-human' ingenuity can be justified. Most uncertain is how AI 'autonomously generates' inventions, and in what way 'AI-generated' inventions differ from inventions developed with the aid of AI. This article would further try to examine the subject topic in detail.

Recent Developments

On 21 December 2019, the European Patent Office (EPO) declared its rebuttal to examine two patent applications, entitling an AI system DABUS as the inventor. The formal ground cited for this denial of examination was the failure to fulfil the requirement of the European Patent Convention that 'an inventor designated in the application has to be a human being, not the machine'2. For the same patent application the United States Patent and Trademark Office (USPTO) asserted the 'Application Data Sheet' with the patent application did not identify the inventor by its legal name. However, the South African Patent Office's verdict to allow the patent has come as disbelief to the global community, moreover the whole incident has riled up the intellectual property experts.

Soon before all this, the World Intellectual Property Organization (WIPO) issued "a call for comments" on the Impacts of AI on IP policy raising, among the member nations, the query of how patent law and policy should respond to inventions 'autonomously generated by AI'.3 The aforementioned initiative was headed by a request for comments by the U.S. Patent and Trademark Office (USPTO) addressing similar issues.4

Necessity Of Patent Protection For Ai-Generated Inventions

Proponents believe that Patent protection ought to be offered for AI/ML produced innovations since it will incentivize innovation. The possibility of securing a patent will not directly encourage an AI module, although it will inspire a few individuals to develop, own, and use AI. Permitting patents on AI-made inventions, therefore, will stimulate the advancement of ingenious AI, which will eventually result in more innovation for society and for its betterment. For instance, a broad-spectrum powerful antibiotic developed with the help of virtual intelligence could be highly useful to mitigate deaths due to infections around the world.

One of the doctrines of intellectual property rights is to stimulate the revelation of knowledge and the commercialization of socially useful products. Patents for AI-produced innovations would help us to accomplish these goals as well as any other patents. On the contrary, failure to grant protection for inventions generated by AI would mean that, in the time to come, enterprises would not use AI to discover and invent, therefore limiting the effectiveness of these powerful inventions tool even when it is more effective than humans in solving certain problems.

Permitting a human being to register as an inventor for an AI-generated invention would not be unfair to an AI since it is that person himself or herself creating the AI module and training the module to carry out such an intricate piece of work, moreover AI would not have any vested interest in being acknowledged.

Cons Granting Patents To Ai-Generated Inventions

It is often argued that, where an AI is engaged, chances are that inventing things turn out to be faster, relieved of human biases and possibly cheaper. The ease of inventing brought by AI may lead to an increase in patenting activity, which might in turn lead to low-quality patents, patent flooding and patent trolling. Additionally permitting individuals to take recognition for the job they have not done would diminish human inventorship. It would put the work of somebody who simply asks an AI to unravel a problem on an equal footing with someone who is inventing something new.

Conclusion

In the current situation, IPR policies are needed to be revisited and updated in order to deal with AI-created inventions. At present, ingenious AI may be an inconsequential part of innovation in economic terms. Although AI is enhancing tremendously, human scientists are not able to keep up the pace. Even in the short-to-medium term, this means that innovative AI may turn out to be a significant part of research and development and when it does, it will be extremely challenging if the IPR policies lack concise directions on whether AI-generated inventions can be protected, who, or what, should be listed as an inventor, and who owns these inventions and related patents. Out of the obligations for patentability, the inventive steps or non-obviousness is the most complicated to evaluate, both in theory and in practice.

But the requirement is central to defining an invention, and it is of utmost importance in instances where society and technology are rapidly changing. When it comes to AI-generated inventions, it is thus in the requirement for patentability of inventive step or non-obviousness that most problems lie.

While an invention might be non-obvious to a skilled person, that same invention might become obvious when seen through the lenses of a skilled person who can use a similar AI system to generate it.

Footnotes

1. Nature (https://doi.org/ggm2p4; 2020) last visited 30-06-2022.

2. EPO, 'EPO refuses DABUS patent applications designating a machine inventor' (EPO, 20 December 2019) (https://www.epo.org/news-issues/news/2019/20191220.html) accessed 3 March 2020. See also EPO, 'Summary of the relevant facts and submissions' (27 January 2020) (https://register.epo.org/application?documentId=E4B63OBI2076498&number=EP18275174&lng=en&npl=false) (EPO, 27 January 2020) accessed 3 March 2020.

3. 'WIPO Conversation on IP' (n 2) paras 6-11.

4. USPTO, 'Request for Comments on Patenting Artificial Intelligence Inventions' Federal Register, Vol 84, No 166, 44889 (Federal Register, 22 August 2019) (https://www.federalregister.gov/documents/2019/08/27/2019-18443/request-for-comments-on-patenting-artificial-intelligence-inventions) accessed 3 March 2020.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.