Abstract
Artificial Intelligence, as its arrival into the world of humans, is changing various industries and challenging IPR developed over time. This article looks at how IPR is changing with the development of AI where ownership, authorship, and creativity are the prime concerns. Technologies linked with AI are altering existing laws and seem to call for a review of copyrights, patents, and trademarks. Selection of works may have ambiguous issues of authorship, originality, etc., copyright law faces tough challenges. The Indian Copyright Act, 1957 provides some stipulations but is far from addressing the AI-specific issues. Similarly, under the Indian Patent Act, 1970, there aren’t any clear provisions about the inventions by AI. Patentability in such works becomes questionable. In the cases where AI systems play a role in creating trademarks, ownership and registration issues may arise due to the same. The article calls attention to notable cases, like Diamond v. Diehr; Waymo LLC v. Uber Technologies; and more that put in limelight the legal as well as practical predicaments Wilkins finds overwhelming need for IP law updates-ones that take into account the unique ability of AI and find a way to delineate its legal status and ownership while ensuring accountability or responsibility for such. Meaning, therefore, in protection of trade secrets for AI innovations, it is important that effective enforcement mechanisms and clarity be provided in order to foster innovation and protect the rights of creators in the age of AI.
Keywords: Intellectual Property, Artificial Intelligence, Copyright, Innovation, IP Rights, Challenges
Introduction
Sometime early in this century, the intelligence of machines will exceed that of humans. Within a quarter of a century, machines will exhibit the full range of human intellect, emotions, and skills, ranging from musical and other creative aptitudes to physical movement. They will claim to have feelings and, unlike today’s virtual personalities, will be very convincing when they tell us so. – Ray Kurzweil (2008)
Artificial intelligence (AI) is here, and it has caused considerable disruptions in healthcare, banking, and other industries. Intellectual property rights (IPR) are in a particularly thorny spot among the many realms impacted by AI. AI, animated and altered software developments biomutate conventional notions of ownership, authorship & invention, challenging extant IP regimes. In this article, here we are going to have a look at the rapidly transforming habitat of intellectual property rights in the fringes of the movement of artificial intelligence.
What is meaning of Artificial Intelligence (AI)?
Artificial Intelligence as the theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages. – Oxford Dictionary
The branch of computer science involving the creation of the devices or systems that can perform the operations normally requiring human intelligence. These tasks, such as learning, problem-solving, understanding natural language, recognizing patterns, among others, are means through which artificial intelligence, by use of data and algorithms, enables machines to be able to make judgments and learn from mistakes together with adapting to new inputs. It’s employed in banking, transportation, healthcare, and from a vast array of fields of endeavour.
The term “intelligence” in artificial intelligence makes an assumption that these machines have capability to decide or even act in a reasoned manner of the information at hand. Since we as human are capable of making conclusions on our own intelligence through looking at information from our immediate surroundings, humans have been termed otherwise as intelligent beings.
The brain of humans is remarkable in that it can learn new things, digest billions of pieces of knowledge per moment, solve issues, apply logic, and a lot more. AI simulates this intelligence through machine learning. After being given information, they are instructed to behave like humans. A machine can be considered intelligent when it is able to act independently and justify those actions at the same time.
In other word Artificial Intelligence can be stated as the ability of machines (computer systems) to carry out actions that usually require human intelligence, such as making decisions and solving complicated issues.
Impact of Artificial Intelligence on Intellectual Property Rights
- Copyright and Artificial Intelligence
Intellectual property rights, such as copyright, are protected in India by the Copyright Act, 1957. The act consists of numerous protections of the author’s exclusive rights to their creative works of literature, drama, music, and the art, as well as recordings of audio and cinematographic films. Some issues related to copyright which need to be solved in the case of works created by AI.
One of the crucial issues is the authorship and ownership of work generated by Artificial Intelligence (AI). In general, by rule of the Copyright Act, the creator of a work will be the author. This way in the case of work generated by AI, the authorship might not be anyways clear and whoever creates it, may not even be the author much less is the intelligence behind AI. Additionally, the issue of ownership is blurry as it might not be evident who is the owner of the copyright in an AI generated work. However, the Copyright Act does expressly provide a special provision u/s 2(d)(vi) that specifies that
“In relation to any literary, dramatic, musical or artistic work which is computer-generated, the person who causes the work to be created” shall be the ‘author’,it does not definitely address ownership right in relation to work generated by AI.
However, calming authorship is one thing and copyrightability is another. The author submits that mere manifestation of authorship alone is not sufficient for copyrightability, which requires the work to be original in addition to being created by an author. After the Indian apex court decision in Eastern Book Company v. D B Modak and Anr. that a certain ‘minimal level of creativity’ is required for a work to be eligible for protection under the copyright act, and a catena of decisions that followed thereafter it is now a settle principle that copyrightability in derivative works depends also on application of skill and judgment in addition to use of labour and capital. Even if the authorship may be assigned to an individual for outputs generated by AI, it does not get one fair if then such outputs generated by AI does not pass muster of the originality required to be met for enjoying copyright protection.
As a whole this act offers some direction on handling computer-generated works including those generated by AI, but a great many of copyright issues related to AI remain unresolved in India. As AI technology progresses, it is crucial for copyright legislation to develop and adjust to safeguard creator’s rights while promoting innovation and technological advancement.
- Patent and Artificial Intelligence
The Patent Act, 1970 in India which lays down the criteria for patentability and the patentability of AI generated invention is determined by this act. The Act provides that an invention should be novel, non-obvious and have industrial applicability. However, there is no specific provision in the Act that deals with Computer Related Inventions (CRI) and Artificial Intelligence (AI).
Invention generated wholly by Artificial Intelligence and Computer Related Invention without the involvement of human mind are not eligible getting patent protection as they do not meet the criteria of being invented by an actual person. However, in case Ferid Allani v. Union of India & Ors, Delhi High Court held that inventions based on computer programs where in today’s digital world when most inventions are based on computer programs, it would be retrograde to argue that all such inventions would not be patentable.
Additionally, the Indian Patent Office has issued guidelines for examination of computer-related inventions which may include AI generated inventions. These guidelines provide that use of AI and machine learning algorithms in an invention may be considered to contribute to inventive step and industrial applicability provided these are not merely being used for automation and/or computerisation of known tasks. Further the guidelines also state that the invention should include description disclosing how the invention provides technical advancement over existing related art by using AI or machine learning.
- Trademark and Artificial Intelligence
Artificial Intelligence is increasingly used in various industries across India such as marketing, advertising, and brand management. Consequently, Artificial Intelligence systems can either record trademarks or assist in the trademark development process. However, the Indian Trademark Act of 1999 does not explicitly address the issue of AI generated trademarks.
As per Indian trademark law, trademark is further describing u/s 2(1)(e) ‘certification trade mark’ that is actually being used or proposed to be used in relation to goods or services with the intention to denote a link between such goods or services and the owner of the mark. Trademark registration is carried out through the Trademarks Registry and the Registry involves examination of marks for registration. Artificial Intelligence is a useful tool in providing quick and correct result of trademark registration but one should remember that under the trademark act of India no trademark having goodwill or possibility of likely deception with other trademarks cannot be registered. Therefore, for any trademarks that are produced and developed by an AI system, they have to meet these qualifications if the trademarks are to be registered. In case where trademark is generated by using Artificial Intelligence their complexity increases in regard to ownership, in general the actual person who generated trademark using Artificial Intelligence would be deemed any rights concerning the trademark to another person or legal entity unless there is a special licensing agreement that assigns the said rights i.e. ownership to another party.
In conclusion, even though there are steady advancements in the implementation of AI technologies in the course of trademark registration, it in is evident that any trademarks created by an Artificial Intelligence need to fulfil the legal provisions that trademark developed by human beings have to fulfil in order to be registered and protected. under Indian law.
Challenges associate with considering Artificial Intelligence under Intellectual Property
There are several challenges associated with artificial intelligence (AI) and intellectual property rights (IPR) in India. Some of these challenges include:
- There are arguments as to whether developers or users of the AI should own its intellectual property for any material that it creates. Present laws give copyright only for works that are a product of skills applied by a human. According to patent law, an inventor has to be a human being. The law is not clear and silent on the matter when AI assists or makes work.
- It itself outlines the need for clear legal status when AI makes decisions which are harmful, especially in cases where AI works autonomously. It should be well defined law. Legislation should be in a place where humans can control the decision of AI. So, creators should be responsible and they should pay the penalty for AI’s mistakes.
- Since Artificial Intelligence is essentially a creator, then obviously intellectual property laws should be adapted. One of the proposals is that users must agree broadly with developers to define and cross a deal on the ownership of AI creations. IP laws should be updated in such a way that it has laws regulating works created solely by AI.
- Even if a business is able to secure the IP protection for its AI-related inventions, the enforceability of those rights is equally problematic in India’s legal system. Delays, besides backlogging of cases, are the vices the Indian legal system is prone to, and these work contrary to speedy enforcement of IP rights by businesses.
- It may be stated that the IP laws existing in India are still developing and hence have certain grey areas. For example, protection of trade secrets relating to AI is not provided by any specific law. Thereby, it becomes arduous for businesses to protect their trade secrets relating to AI.
Famous cases based on Intellectual Property and Artificial Intelligence
It is then imperative to accept that with the advancement of modern society, there is no part of the society that is immune from the effects of technology. The penetration of technology into the society blurs the real world and the cyber world making the latter look real painting a picture that conceals the real world with the cyber world.
There have been several notable cases involving Intellectual Property and Artificial Intelligence, including the following:
- One of the most striking cases involving the patenting of machine learning algorithms involves the case of Diamond v. Diehr in the United States in 1981. In this case, the patent application sought was for a process for using a computer to calculate with precision the correct curing time for rubber. It involved feeding in temperature readings into a computer and then using software within the computer to compute the optimum curing time for the rubber. The US Patent and Trademark Office ruled against the application on the basis that the concept was an abstract idea and not an invention. However, the United States Court of Appeals for the Federal Circuit reversed the decision, finding the use of a computer to make the calculations made the process patentable. CAFC determined that the process was not an abstract idea but an application of a mathematical formula in a specific and practical application. The court also used to do the calculations as a major part of the patentable invention. This principle built from this case, that software and algorithms can be patented if used in a practical application rather than an abstract idea. The case also emphasized the importance of the use of a computer in making the process patentable.
It was a major breakthrough in the history of machine learning algorithm patents because it established precedence and paved the way for the rest to be patented for software and algorithms applied in practical use.
- Recently, cases have emerged where the question of who owns the copyright to the artwork generated by an AI arose. The Indian Copyright Office had one such significant case in 2020. An AI system, named ‘RAGHAV‘, attempted to get an artwork titled ‘Suryast’ registered for copyright protection. The first rejection was because there was no human author listed for the painting. Another attempt at protection was sought and allowed with a natural person this time listed as a co-author along with ‘RAGHAV.’ Then a notice for withdrawal was issued, this time lifting the legal status of ‘RAGHAV’ for clarity on if under the Copyright Act, AI would also be considered an artist or otherwise.
- Waymo LLC v. Uber Technologies, Inc. is a dramatic example of trade secret theft by a former employee. Waymo is a subsidiary of Alphabet Inc. and developed self-driving technology. Uber is in the business of ride sharing and acquired a self-driving trucking company called Otto founded by Anthony Levandowski, a former employee of Waymo. In 2017, Waymo sued Uber based on allegations that Levandowski stole more than 14,000 confidential files relating to Waymo’s self-driving technology. It claimed that Levandowski had downloaded onto a hard drive certain files originating from Waymo’s servers and that Uber had later hired him as a means to get access to Waymo’s technology. 2018 saw the settlement of the lawsuit by Uber for 245 million and consented not to use information marked confidential on Waymo in its self-drive technology.
Conclusion
The emergence of artificial intelligence has brought tremendous and radical changes in many industries such as healthcare, banking, and so on, thus raising the many complex issues with regard to IP rights. AI is creating new ways of authorship and ownership and creativity which gives the need for taking a fresh look at the existing intellectual property rights regimes as it becomes developed. The person who causes the work to be made is identified as the author by the Indian Copyright Act, 1957 and some guidance on this issue of computer-generated works is provided in so far as copyright is concerned. When it is AI-generated, questions more particularly about what constitutes originality and creativity in copyright protection remain. Similarly, problems exist while determining the patentability of AI-created inventions since the Indian Patent Act of 1970 does not contain any express provisions relating thereto. Although the contributions by AI and machine learning can be taken into account under inventive step or industrial applicability while examining the guidance issued by the Indian Patent Office, these rules primarily relate to inventions that use AI and not ones that are absolutely AI-created. Uncertainties surrounding trademark law flow from AI’s ability to generate trademarks also because of doubt as to ownership and requirements of registration.
Therefore, Intellectual Property needs to be updated and clarified a lot towards such issues but also for taking into account the unusual aptitude of the AI. Laying down IP rules that would recognize AI as a creator; assuring clear culpability for actions made by autonomous AI systems; and defining the legal status and ownership of works technologies. Additionally, procedural backlogs and case backlogs make it difficult for IP rights to be enforced in the Indian legal system, which in essence, makes it difficult for companies to protect their AI-related ideas. Also, India’s rules of law relating to IP rights are still in their infancy and there are also some areas where grey areas remain, undefined special legal protection for AI-related trade secrets being one among them. IP laws must catch up with AI reality in such a way that spares no room for artists’ rights and innovation incentives. This means that there have to be clear standards in AI-generated work ownership, originality, and patentability while having effective methods to enforce them.
Author:
Uditya Kumar
The ICFAI University, Dehradun