Author – By Chitwandeep Kaur & Shivam Nayyar (Intern)
The notion of Artificial Intelligence is not alien to us anymore. AI is becoming part of our day-to-day technology at a rapid pace. We may or may not have a clear understanding of what exactly AI is and how it works, but are using a number of AI on a daily basis intentionally or unintentionally. Some of the most common forms of AI are Alexa- a voice assistance AI, Navigation Maps, Chatbots etc.
Various scientists have given their own definition of AI. Some define AI as a branch of computer science’s machine learning that aims to create “intelligent machines”.
Authors Rich and Knight (1994) and Stuart (1996) described artificial intelligence as the capability of machines to perform the same tasks as human beings.
In simple words we can say AI is the ability of a machine to do tasks that are usually done by humans using their intelligence.
AI works using data or algorithms that grants the machine the ability to learn automatically from patterns or features in the data. AI is a broad field of study that involves many complex technologies and subfields. The above definition and working is given to provide general overview to a non-technical person.
IPR and AI
The AI interacts with IPR in several aspects. The integration of AI has eliminated the traditional practices of IPR. The advancements in AI have made it capable to generate IP assets for companies, buy products for consumers and many more. AI interacts with Trademark, Copyright and Patent Law in various ways.
Interaction of AI with Trademark Law
Trademark is a sign used by companies to distinguish their products and services from their competitors. The unique features of a Trademark guides consumers to its source. But with the involvement of AI in aspects such as e-commerce reduces the requirement of human cognizance. We are all familiar with Alexa- a voice assistance AI which receives commands from human voice and acts accordingly. If we want to buy a certain brand’s product on E-commerce, we can command Alexa to do our task. Tasks done by AI may be partially or sometimes fully automated. This reduction in human cognizance makes several traditional trademark concepts such as “likelihood of confusion”, “concept of average human” insignificant.
Now with the above example it is clear that involvement of AI in e-commerce has reduced the human cognitive process in selecting or placing order of goods. Let’s try to understand the Trademark Notions and their interplay with AI.
The key objective of Trademark Law is to protect the rights of the owner and the consumer from counterfeit products. The use of AI in online shopping influences the purchasing power of human beings and reduces the possibility of purchasing counterfeit goods. Since the AI are programmed in such a way that they do not require Trademark to identify the source, whereas they can directly identify the product by analyzing data.
The appliance of AI ousts various Trademark function which relates to human psychology:
Likelihood of confusion: It is a notion of trademark law, which exists when the marks are so similar and goods are so related that the consumer gets confused. It is one of the most persuasive bases of infringement and serves as a tool to determine similarity of marks.
Average consumer with imperfect recollection: In this notion the perspective of fictious man of ‘average memory but with his usual imperfections’ is used to guide the decision of the court in case of two similar Trademarks.
Trademarks engage with consumers on a psychological level, contributing to positive or negative associations, hence, traditional notions stated above are emphasized. The AI applications influencing purchasing decisions and eliminating the human cognitive process makes these notions irrelevant. However, there are few AI systems which are not fully automated and require human supervision for example Recommender System.
Thus, it can be concluded that the AI does not entirely challenge the Trademark Law functions, most of the applications only provide recommendation, support or assistance but still require human participation. However, some applications which are fully autonomous may eliminate the Trademark Law functions.
Interaction of AI with Patent Law:
AI enabled machines are now capable of performing innovations based on their learnings and intelligence. These inventions generally require some human assistance or but it may sometimes be completely on their own. This new development seems magnificent but it comes along with unfamiliar threats and challenges to the current legal mechanism.
The issues like whether AI inventions are patentable or not? If they are patentable, who is the owner of inventions?
Patentability: First Challenge regarding patentability of AI invention is whether the invention by AI passes the patentability criteria. A crucial factor for any invention to be granted a patent is, whether or not it can pass the patentability criteria. In most countries including India, a patent needs to possess novelty, an inventive step, and capability of industrial application.
To ascertain novelty of an invention, the inventor is required to differentiate it from the already existing creations. To establish this differentiation, it is tested whether or not the invention requires a creative mental component. If the invention is easily anticipated from prior art and does not require much mental labor then it is not novel. AI systems have an extensive database of prior arts and determining the novelty of its inventions is difficult. As far as inventive steps are concerned it is established on the concept of innovation and non-obviousness. If novelty is not established, achieving inventive step is much more difficult.
The legal mechanism still does not consider AI as an inventor or patent holder.
Section 2(y) of the Indian Patent Act defines “true and first inventor”, the word “person” used in the section to define inventor and further the intention behind the use of “person” makes it clear that the Act recognizes only humans as inventors. Further section 6 of the said Act states who can apply for a patent. Reading both the sections together gives us the perspective that only humans can apply for patents.
Similarly, the US Patent Act, also does not recognize any other entity other than human as inventor.
Apart from the current legal mechanism, the objective behind the concept of Patent Law has the same viewpoint. The objective to grant patents was to give reward to the inventors in form of monopoly or other monetary benefits to motivate them to carry on inventions. But this concept seems to be ineffective for AI as they don’t have ability to feel emotions such as motivation as of now.
Interaction of AI with Copyright Law
The latest AI machines are capable of creating artistic works such as music, painting, poetry etc. Sometimes such creations do not require any human assistance. With such advancements it has become difficult for the authorities to determine ownership copyright over creation by AI. Further, they are having trouble identifying the possible infringement by AI.
The Copyright Laws of various countries recognize only humans as the author of any creative work. In the case of Naruto v Slater(Monkey-Selfie) wherein a Monkey took his selfie on its own using Photographers camera. In this case there was a dilemma as to the ownership of that photograph. The US Court held that animals cannot be granted copyright protection as only humans are entitled to it.
In the current situation the Authorities have a similar viewpoint regarding the copyright ownership of AI machines. The AI does not get the ownership rights for its creations.
In India also, humans are only considered as Authors. According to Section 2(d) of the Copyright Act, 1957 by providing that “in relation to any literary, dramatic, musical or artistic work which is computer-generated, the person who causes the work to be created.”
Although, machines like computer systems can be used to generate work, the only thing necessary is human intervention or creativity in the creation of work to make it copyrightable.
In the case Amarnath Sehgal Vs. Union of India, the Court again put emphasis on the fact that AI should be excluded from the ambit of authorship and only humans can be an author.
The legal statutes and Judicial Precedents make it obvious that AI can be the owner of its creation. There is one more issue that is, if the AI infringes someone’s right then who would be liable for that?
From the above discussion it is clear that the existing IP laws are incapable to align with AI advancements. The businesses nowadays are investing in AI technology more than ever before. Therefore, the modification in IPR laws and its effective implementation is necessary to overcome these challenges.