This piece has been authored by Milan Prajapati and Komal Jain, 4th Year Law Student, Institute of Law, Nirma University
Introduction
Generative AI creates different types of content using algorithms and analyzing the existing data. It has the capability to even surpass human-generated work and creativity. But the question that arises is who should own the copyright of the work created by AI and who should be held responsible in case of any violation. To answer these questions, we need to look at the existing IPR framework and bring necessary changes to the legislation to ensure fair protection.
The technological transformation in the 21st century that we are experiencing has forced us to reconsider how computers and the creative process interact. The fast advancement of machine learning software, a kind of artificial intelligence that creates self-governing systems with the ability to learn without explicit human programming, is the foundation of this revolution.
Content produced by artificial intelligence has had a big impact on copyright laws. In the past, it was believed that computer-generated works’ copyright ownership could not be challenged because the software acted as a pen-and-paper replacement to aid in the creative process and it has nothing to do with human creativity. A creative work is protected by copyright if it is original and as per all the definitions of originality, it is required that the author must be human. Most nations, like Germany and Spain, stipulate that only works created by humans are eligible for copyright protection.
EU AI Act and USPTO Report
The European Parliament has presented a draft Artificial Intelligence Act that, among other things, imposes duties on generative AI systems to adhere to transparency criteria and ensures safeguards against producing illicit content. The Act’s goal is to regulate the safe and open use of AI. The primary foundation of the Act is a classification framework that assesses the potential risk level associated with AI technology in terms of its impact on the well-being, safety, and basic rights of individuals. The system comprises four distinct risk classifications, namely unacceptable, high, limited, and low. The proposed legislation additionally highlights regulations pertaining to AI systems categorized as general-purpose AI, denoting its capacity to be employed for diverse objectives with differing levels of risk. Examples of such technologies encompass huge language model generative AIsystems such as ChatGPT.
A report on the public’s perceptions of artificial intelligence and intellectual property policy was released in the US by the US Patent and Trademark Office (USPTO). The report provides a comprehensive examination of several stakeholder perspectives about the implications of AI on the Intellectual Property domain. It encompasses various aspects such as patent, trademark, copyright, and trade secret policies, along with emerging concerns related to database protection. It exemplifies the agency’s strong dedication to staying abreast of this swiftly evolving and crucial technology, intending to expedite American innovation.
Generative AI Works in the International Domain
There are two possible approaches to addressing authorship of works made by computers: one option is to attribute authorship to the programme creator, while the other alternative is to deny copyright protection for such works altogether. The explicit prohibition of granting copyrights to works created by artificial intelligence has yet to be established.
As per the regulations outlined by the Copyright Office in the United States, it is stipulated that the registration of an original work of authorship is permissible, on the condition that such work was generated by an individual. The aforementioned stance is derived from legal precedents which establish that copyright legislation specifically protects “the products of intellectual effort” that “originate from the imaginative faculties of the intellect,” as shown by the Feist Publications v. Rural Telephone Service Company decision.
The issue of determining copyright ownership for works produced by AI remains enigmatic and challenging, despite governments acknowledging the awarding of copyrights to such works. This is due to the existing legal framework, which mandates that a right holder must possess legal personhood. However, an AI lacks this status until its creator is bestowed with it on behalf of the AI. Nevertheless, a potential loophole arises in this context regarding the allocation of copyright ownership if the AI system is acquired via a purchase.
In the legal matter of Bleistein v. Donaldson Lithographing Co, the Court distinctly distinguished between the creative output of a human being and an artificial creation. In the majority opinion, Justice Holmes articulated the distinctiveness of human personality and established it as a necessary condition for copyright protection.
The legal case of Burrow Gilles Lithographic Co. v. Sarony pertained to the distinction between creative and mechanical employment. The Court deliberated on the potential for conferring copyright protection for a product that is generated by a mechanical device. The court’s ruling, which asserts that simply mechanical employment is inherently lacking in creativity, has limited the extent of its protection.
Indian Perspective
The Copyright Act of India, which was passed in 1957, does not address artificial intelligence-generated works directly and does not recognise AI as an author. For AI works to be eligible for copyright protection, they need to be novel and creative. This is one of the most important requirements for copyright protection because the AI depends on data obtained from pre-existing sources available on the internet as well as data supplied by humans during training, the content that it produces could not match the criteria for originality or innovation.
Originality is a standard that is used to establish whether or not a work is protected by copyright. According to Section 13 of the Indian Copyright Act, the existence of copyright is acknowledged in “original literary, dramatic, musical, and artistic works.” However, the Act does not provide a clear definition of “originality,” thus it is up to the courts to decide whether or not a particular piece of work fulfills this requirement. “Computer-generated work” as per the “Copyright Act” in India was changed to cover computer-generated works, which may include literary, dramatic, musical, or artistic works. These types of works can be created by a computer. The statute was amended to include Section 2(d)(v), which defines the authorship of such works as “the person who causes the work to be created.” This provision was included so that the act could provide a legal definition of authorship.
India’s legal system asserts that individuals who are natural beings are the only ones who can be given credit for writing a copyrightable work. The Indian Copyright Office adds that people have also been guessing and worrying about how to handle apps that contain works made by AI. An application from 2020 said that an artificial intelligence (AI) called RAGHAV was the only author of a piece of art but it was turned down. Despite this, a second application was sent in, and this time, both a person and an artificial intelligence were recognized as co-authors. As a result, the registration was given. This notice says that it’s possible that the registration was given wrongly and without enough thought being given to the issue. Attributing authorship to a person is important for works created by AI to get legal protection in India.
By fostering innovation, entrepreneurship, and economic growth while protecting the rights and interests of its creators and inventors, it may enhance the field of AI. An adaptive IPR policy will be crucial to India’s sustained success in the rapidly changing IP landscape as generative AI continues to change our world.
Conclusion
Significant progress in the field of computing, coupled with the extensive computational capabilities at our disposal, may render the aforementioned difference inconsequential. By endowing machines with the ability to acquire stylistic knowledge from vast collections of data, their potential to imitate human behaviour will continue to improve. The most logical course of action seems to be the attribution of copyright to the individual responsible for enabling the functioning of artificial intelligence.
However, giving copyright protection to AI will violate human labour and talent to produce artistic, literary, cinematographic, and dramatic, etc kinds of artwork. Though the AI solely works on the software and the algorithms built by a human, giving the creator the copyright is still justifiable but not the AI. The increasing prevalence of AI in artistic endeavours is anticipated to introduce heightened complexity. As machines continue to enhance their ability to generate creative works, the line between artwork originating from human creators and those produced by computers is expected to grow more indistinct.