top of page
  • SAINIKITHA.OL

DISPUTES REGARDING THE OWNERSHIP AND LICENSING OF CONTENT GENERATED BY AI

Sainikitha. Ol,

Tamil Nadu Dr. Ambedkar Law University, School of Excellence In Law

DISPUTES REGARDING THE OWNERSHIP AND LICENSING OF CONTENT GENERATED BY AI

Introduction:

The 21st century has come up with plethora of invention and innovation changing the lifestyle of the mass population, one such marvellous creation is Artificial Intelligence marking a tremendous change in the ordinary lifestyle of the people. In today’s world, AI is used by almost everyone from students to tech employees tend to heavily rely on the digital assistance provided by these learning machines. The advent of generative AI as like Chat GPT and DALLE has further simplified the information dissemination process across the globe. These Large Language Models are trained from abundant sources and works which makes them highly efficient and flawless. Using these models users can generate text, video, audio, pictures and so on by simply providing prompts which are instructions. This aspect of AI has attracted unprecedented legal issue as to who owns the ownership rights over such contents generated by these LLM’s since there exists no legal statue or pronouncements explicitly addressing this segment of AI generated contents. But, few provisions in the Indian Copyrights Act, 1954 and few other doctrines like sweat of the brow and significant input test indirectly assist legal luminaries to better address the problems revolving around the ownership of works created by Artificial Intelligence.

Artificial Intelligence and its meaning:

IBM defines Artificial Intelligence as “a technology that enables computers and machine to simulate human intelligence and problem-solving capabilities.” These are made to perform certain designated tasks with highest efficiency and efficacy. Hence, it is even capable of challenging human intelligence but still lacks emotional intelligence. 

Copyrights issues and ownership of AI generated works:

Within a span of 20 years the growth of artificial intelligence has been tremendous and noteworthy. There is a huge shift from simply googling the content from internet to making one with the aid of AI. But this raises myriad issues and questions which is to be addressed by the legislature and judiciary. The first issue is who owns an AI generated work? And what happens when AI infringes on the works of some other persons but this issue is of less relevance given the use of fair use and fair dealing in the copyright law. At present, the Indian Copyright

Act do not have an explicit provision recognising the works created by AI. With the advent of ChatGPT, a text generating AI, this issue has further been exacerbated. The role of human author is essential in order to gain copyrights protection as per the Indian Copyrights regime. But the contents generated by these machine intelligence lacks human authorship. These LLM’s are trained by the companies through billions of data and other available works and hence they are more capable of mimicking human element to the content and they lack creativity aspect and their contents remain less original as they make works completely through the pre-existing contents fed to them for training purpose.

Section 13 of the Copyrights Act clearly states that protection is granted to those works which are original literary, dramatic, music and other artistic works. The act recognises humans as the author and no other creatures. One classic example for this could be traced in the famous ‘monkey selfie’ case where a monkey named Naruto took a selfie using the camera set up by David Slater, a wildlife photographer. An issue was raised whether the monkey could possess ownership over the work and the Court refused to recognise monkey as an author. Similarly, the works of these smart machines might not be qualified enough to granted ownership and protection from infringement as they lack sufficient originality and creativity. Arguments are made regarding this aspect; one group advocates that ownership shall be granted to the AI generated contents and works as they don’t replicate some other’s work but essentially provide  a new expression of thoughts and ideas. But on the contrary, some stakeholders argue that these works lack potential creativity to recognize them as a separate authors and owners. The 1994 amendment to the copyrights act included section 2(d)(v) which recognized computer generated works by adding “the person who causes the work to be created”. AI is neither a natural person nor a juristic person and at present, only the works of natural persons have been recognized by the prevailing laws and statues which needs further scrutiny and developments as AI has become an indispensable part of human life. The Delhi High Court has firmly refused to recognize the ownership of AI generated lists as they lack requisite human labour and intervention.

Granting and recognizing AI as authors is not a simple matter due to multifaceted issues. First, in cases of infringement who will be given remedy or who shall claim the remedy for infringement since AI’s are clearly aren’t capable of claiming remedies on their own and there exists no necessity for the same. Is it viable to grant ownership to the users of such LLM who causes the work to be created by applying suitable prompts? This aspect needs more interpretation and is heavily hinges on the Terms and Policies outlined by the companies and providers of such model. Nevertheless, this too is not feasible as there is no human involvement. Then, shall the company vest the copyrights of all those works is clearly unfair as well as not a feasible solution. For instance, could Microsoft possess ownership rights over all the paintings made with the aid of MS paint?  Sounds unfair, isn’t it? These days, most people rely on AI to bring out better outcomes and hence, this solution won’t be applicable in the broadest sense.

Originality and works of machine intelligence:

The following requisites should be met in order to gain ownership and copyrights,

•      Novelty

•      Originality

•      Must not be available to the public before seeking protection i.e. must not be in public domain

Originality is never been compromised meaning that it is crucial for any work to be original in order to gain protection. It does not concern as to the originality of idea but as to the expression of such ideas. The degree of originality varies from country to country. A fundamental problem in providing ownership to AI creation is lack of originality as these LLM’s are not capable of possessing their own thoughts and ideas to create a work. It is clear that AI lacks originality aspect but contributes for the creation of several works which can be recognised by applying few doctrines.

a) Sweat of the brow – A simple way yet the least applied standard for gaining copyrights over a work would be via the “Sweat of the brow” doctrine. A mere effort added or through diligence a person could gain ownership over a work and this doesn’t apply to the Indian system. This was recognised in the significant case of University of London Press Ltd v. Tutorial Press Ltd where a sufficient amount of effort added to a preexisting work was given protection.  

b) Significant Input Test – This test acknowledges the works created by Gen-AI by integrating the requirement of sufficient human involvement which causes the creation of the work in question. Applicable in the United States where ownerships are granted to AI assisted works by analysing the extent of human involvement applied to the creation of the composition.  

Grant of ownership to AI generated creation – Raghav’s Case:

In 2020, an AI-based tool called Robust Artificially Intelligent Graphics and Art Visualizer (RAGHAV) was deployed by Ankit Sahni to generate ‘Suryast’ and sought copyrights protection and ownership over the painting. At the first instance, an application was made in the name of AI RAGHAV which had been denied ownership and protection by the Indian Copyrights Office. And in the second attempt, application was registered in the name of Mr. Sahni with RAGHAV being the co-author, who contributed for the creation of the artwork. With the second attempt being successful, Sahni made a remarkable history by being the first ever person to be granted copyrights for AI assisted work by the Indian Copyrights Office.

Issues with regard to granting ownership rights:

i. These works lack creativity and originality

ii. By using a single line prompt, a user shall not gain ownership which affects creativity

iii. Ownership rights and protection remains with the author and their heirs for about 60 years from the date of demise of authors. This aspect is undoubtedly not applicable to AI since they have perpetual existence.

iv. Several dilemmas arise in cases of infringement issues. Can AI sue or be sued? For this purpose, the legislature must first attempt to define what an AI is and its legal position.

Conclusion: 

Artificial Intelligence has become an indispensable part of human life and instead of avoiding their potential, the lawmakers must aim to address this ever-growing issue. Different countries have made various provisions and standards to recognise the contributions of AI generated works and creations. US applies Significant input test and Canada adopted a system of authorising AI as co-author and many other countries have embarked on making suitable legislations in order to keep up with the surging advancements made in the tech savvy world. After reviewing, the 161st report of the Parliamentary Standing Committee has suggested and recommended to analyse the rights of Artificial Intelligence pertaining to their relevance with Intellectual Property Law. Therefore, a comprehensive legal framework is essential to address this pressing issue addressing the role of AI in the creation of literary, artistic, music and other works.

References:

 

 

32 views0 comments

コメント


bottom of page