Artificial Intelligence and Intellectual Property In India

Overview of AI and IP

Artificial intelligence is a rapidly evolving piece of technology, which has massive impacts on all industries and economies around the globe. AI, a simulation of human intelligence by a computer system designed to perform tasks typically done by human beings, has in just a few short years evolved beyond what was previously believed possible, and has manifested itself in even creating new works of “Intellectual Property”, such as art, literature, music, etc. As Josh Antonuccio, an associate professor at Ohio University School of Media and Arts stated: “It is a tool far beyond the scope of anything we have seen in human innovation at this point”. AI redefines the concept of creativity and innovation while leading certain intriguing questions, such as whether these “IP” made by AI can even be considered as intellectual property or whether these “innovations” belong to the AI or the programmer of the tool, or if these innovations can even be considered original to begin with?

One of the most widely utilised aspects of AI has been voice manipulation and Deep fakes. AI has increasingly been used to create literary, artistic and musical works, circulating all over social media, and it has become the gateway for the general public to the domain of AI. There have been several videos and sound bites circulating online using the AI versions of celebrities singing songs, dancing, giving a speech, etc. Most popularly, the musical artist Drake, used the deceased artist, Tupac’s voice, in his song “Taylor Made Freestyle”. AI uses available data, such as in this case sound bites of Tupac’s voice from interviews, old songs and public sightings, and uses it to capture the various eccentricities in the voice, such as the depth, tone and enunciation. The AI is then fed commands, in this case lyrics and music bytes, to mash the voice of the artist and the musical commands, to create an “innovative” song. The basic idea behind AI is that it is fed “problems”, and based on the algorithms and data available to it; it comes up with new “solutions”, it not only assists the users but also creates new works. Similarly in deepfakes, AI manipulates the visuals and audio data to create a hyper-realistic but fabricated recreation of a person, doing whatever the creator wishes to see them do. This gross violation of public personas extends beyond artistic expression and opens a minefield of misuse. In India, most notably videos of the popular TV personality Rajat Sharma were seen circulating online where he was seen spreading misinformation, damaging his reputation as a credible journalist. These videos were revealed to be AI-generated deepfakes. The journalist sought a permanent injunction, contending the wrongful use of AI infringed IP and personality rights. However, due to the absence of a legal mechanism for the regulation of deepfakes and AI, he had to file a PIL. These uncanny and exact replications of celebrity voices and visuals lead to a world of problems not just from a privacy point of view, but in general IP as well. With AI, innovation and creativity is no longer a talent, a person can get a plethora of new works just by typing in simple commands, often using celebrities to gain traction, diluting the actual art form, and using celebrity replications to advocate for whatever harmful messaging they wish to perpetuate. In a country where celebrity worship is the norm, where video and audio media are the choice for information consumption, and the combination of the two has an unfathomable influence and unbreakable hold on the general public, the misuse of AI is a catastrophe waiting to happen.

Copyright is a legal entitlement which is bestowed upon the authors of creative works, which includes music producers and directors. Currently, the law of Indian IP does not recognise AI as an entity which can ask for protection. As per the definition of author under the Copyright Act, any person who causes work to be generated by consumers is considered to be the actual author. Therefore, under this definition, the person who created the prompt for the AI shall be given the title of the owner. This creates an issue, as the creators of the machine, and the actual makers of the music are denied the fruits of their labour, particularly in cases where no consent was provided by a person, but due to the freely available data, AI was able to copy their likeness and style.

In India, in copyright, originality is given priority; however, the term original works lacks precision. The courts examine the idea and the execution, the connection and the author’s skill and labour, before deeming it original. AI lacks these conventional components of idea, creativity, and intuition; it functions on a large-scale database, performing tasks on predetermined algorithms. Even in the case of splendid replications of a human likeness, from voice to image, the works of AI still lack the conventional “creativity component”, which in all definitions set by law would not make it eligible for authorship.

The ethical dilemma arising from works of AI is that it strips the powers of the person whose likeness it is using, it zaps the creative process which is usually put forth in works of creativity, it decreases the labour and often leaves a product bearing the markings of previously tried and tested works, leaving no room for originality or innovation. AI is a system procuring data, if previous data is the only metric used to justify the existence of art; we are only left with art made by the “formula” for fame. By the process used, the art produced is purely a different iteration of what “works”, a big name, flashy sounds, popular beats, buzz words, and just enough bold imagery to capture the viewer’s depleting attention.

The popular notion is to assign AI a separate identity, as it will not only deter those looking to capitalise over easy works of AI, it will also give recognition to the works of AI. However, this may prove difficult as in the case of granting the author rights, the moral rights provided under Section 57 of the Act are nearly impossible to enforce. AI cannot be held accountable for infringement of personality likes, or any objectionable content it may produce, as it lacks the capacity for moral judgement. Providing authorship to AI, further gives a loophole to the authorities feeding commands to AI, to get away with producing morally questionable pieces of media, and giving no outlet for the persons whose imagery and likeness have been misused. Providing AI with a separate identity may lead to further misuse and no accountability for those actually utilising the AI to create harmful forms of media.

Currently, the Indian legal system lacks any form of substantial piece of legislation in the works of AI. It is a given that works of art entirely generated by AI cannot be given any form of protection, however, the artistry used in prompting the AI, to elaborate on media, using the unlimited potential to create something genuinely interesting and creative, should be given its due credit. Further, there should be accountability in the field of personality rights; previously in man-made “remixes” of existing IP, obtaining consent was a compulsory step. With AI, this aspect has been eradicated. Therefore, there should be legislation providing a system of checks and balances, leading to greater accountability and less misuse. India could have a separate law dealing with AI, as has been done in several other countries. It is a fact that AI is evolving at a rapid pace, and its impact is here to stay, its ever-growing advances have launched us into unknown territory highlighting the lacunas in the legal system, particularly in the field of IP. The questions posed by the existence of AI systems can only be answered by a separate law; the issues can only be solved by exploring and proposing thoughtful solutions.

Author: Aadya Tewari, in case of any queries please contact/write back to us at support@ipandlegalfilings.com or IP & Legal Filing