Artificial Intelligence (AI) stands as one of the most revolutionary and controversial inventions of the mid-twentieth century, notable for its capacity to learn from experience and operate independently of human intervention. Its significance lies in its ability to function autonomously without human assistance. Although the concept of artificial intelligence has existed for over 60 years, its extensive application has only become prevalent in recent times.
We see the potential of AI in our daily lives right from playing games, automated suggestions on the internet, face recognition, self-driving cars, etc. It has been used in various fields including healthcare, legal field, traffic control systems, and many other fields. Owing to autonomous features and the ability to learn without the assistance of humans, the existing sectors of IP are facing a range of difficulties, particularly in the realm of copyright.
Existing copyright laws are sufficient to protect human-created content. However, AI-generated works pose new challenges. There is no clear legal perspective on AI, making it difficult to determine if AI-generated works can be protected by copyright. A primary issue is whether such works should be protected under current laws. If AI-generated works are not protected, it could violate competition laws in various countries. Therefore, to maintain market competition, AI-generated works should be protected under copyright law.
The World Intellectual Property Organization (WIPO) has recently addressed this issue by holding the WIPO Conversation on Intellectual Property and Artificial Intelligence, Second Session, 2020. This session highlighted the need for protecting AI-generated works[1]. This international initiative shows the global concern and efforts to find the best approach for recognizing and protecting AI-generated works under intellectual property law. In this context, protecting AI-generated products legally would preserve competitive spirit and recognize human ingenuity.
Guadamuz, 2017 stated that the rise of AI in art blurs distinctions between human and machine creation, raising questions about copyright and protection. Granting copyright to AI facilitators ensures continued technological investment.
AI significantly contributes to the development of creative works, yet Indian copyright laws appear ineffective in this context. There is on-going debate about whether AI-generated works meet the fundamental requirements for copyright protection under Indian law. Key issues include identifying the author, whether the author is the AI owner, the programmer, or the AI itself, and whether AI can be treated as a co-author or held accountable for infringement. The complexities in defining AI legally, given its independent operation and lack of transparency, further complicate this matter. This study aims to address these issues and propose appropriate solutions.
Artificial Intelligence
Artificial intelligence is “a technical and scientific field devoted to the engineered system that generates outputs such as content, forecasts, recommendations or decisions for a given set of human-defined objectives”.[2]
Artificial intelligence or AI refers to computer systems that have or possess human intelligence. These processes include learning, reasoning, problem-solving, perception, and understanding language. AI systems can analyze data, make decisions, and perform tasks that typically require human intelligence, using algorithms and machine learning techniques. To execute jobs that would otherwise need human intellect, AI does not require the physical appearance as like a robot. E-mail classification, Gmail’s spam filter, Google Translate, Google search engine and Amazon’s recommendation system are all examples of AI algorithms in operation.[3]
Machine learning is the ability to learn without being explicitly programmed. It entails analysing data with algorithms, learning from it, and generating a determination or prediction as a consequence. Deep learning is a branch of machine learning that mimics the intricate decision-making capabilities of the human brain using multi-layered neural networks, or deep neural networks.[4]
Intellectual Property Rights
The rights that people are granted over their creative works are known as intellectual property rights. For a set amount of time, they often grant the creator the sole right to use his or her creation.[5]
IPRs are categorized[6] as follows based on the kind of innovation and creativity of the human intellect and their application:
- Copyright: Authors are granted the exclusive or monopolistic right to their literary, dramatic, and creative creations. It is the copyright holder’s right that forbid others from duplicating or replicating his work. [7]
- Patent: It is a person’s exclusive right to use their innovation for a set amount of time, awarded by the patent office. Only novel, creative ideas that have the potential for industrial use are eligible for patent protection.
- Trademarks: A trademark can be a device, brand, label, name, signature, word, alphabet, shape of goods, packaging, or combination of colours. It serves two purposes: first, it safeguards the public from misunderstanding by clearly identifying the product’s source or origin about other comparable products; second, it safeguards the trade and business of the trademark owner as well as any goodwill associated with it.[8]
- Geographic Indication (GI): They promote the products of producers in a specific area and identify the particular territory from which the goods come; it can also be a region or locality within that territory[9]. It indicates to consumers that given quality, reputation, or other characteristics of these goods are fundamentally attributable to their geographic origin.
- Industrial Designs: Design refers to the characteristics of any object applied by an industrial process. Since industrial design makes an object more appealing and increases its commercial value, it must be protected by law. Legal protection against imitation is necessary for a design’s creative originality. To achieve this, India passed the Design Act, 2000, which permits the registration of designs and offers the owner of the design remedies against proliferation.[10]
- Trade Secrets: A trade, business, or profession’s trade secrets are its exclusive knowledge and techniques. Any knowledge that may be utilized to run a business and provide it with financial benefits over rivals is considered a trade secret. Unauthorized use of such data is protected without registration and is considered an unfair practice and amounts to a breach of trade secrets.
Protection of AI-generated works under Copyright Laws
Works that are fully produced by artificial intelligence systems without direct human involvement are referred to as AI-generated works. These works can be literary, musical, or other creative productions produced independently by AI algorithms. However, even though artificial intelligence tools and software are used in the creation of AI-assisted work, human creators continue to have a significant amount of input and control. In these situations, artificial intelligence (AI) technology either supports or improves the creative process, but it does not fully replace human input.
In the context of intellectual property rights, the distinction between AI-generated and AI-assisted works is significant because it can affect matters like ownership, authorship, and copyrights. While AI-assisted works usually maintain clearer attribution to human creators with AI tools acting as aids rather than sole creators, AI-generated works may raise legal questions regarding who owns the creations.
The AI-generated piece of art “Portrait of Edmond Belamy” can be taken into account[11]. It sold for 432,500 dollars. Since this painting is an original work of AI and possesses unique qualities, copyright should also be applied to it. The production of these kinds of works encourages intellectuals to speak out in favour of granting AI authorship rights and to investigate the feasibility of changing the existing IP legal framework. In “The Monkey Selfie Case,”[12] a San Francisco court ruled that an AI or non-human author should not be granted copyright protection for a selfie that a macaque monkey had taken.
Similarly, the UK Intellectual Property Office refused the patent applications listing DABUS as the inventor. The court emphasised that an ‘inventor’ must be a person. It mentioned that in order for the inventor to be held legally responsible for the patent or the patent application, inventor ship needs to be ascribed to a person with legal personality. Recognizing an AI as the inventor would be intrinsically incorrect, as they lack legal personality.[13]
In order to receive copyright protection, the work must satisfy the following requirements based on a number of copyright law provisions:
1) The work must be unique;
this can be verified by meeting two additional requirements:
Work needs to be:
- independently created;
- creative; and
- fixed/idea-expression. Contrast
A nation’s competition laws would be broken if the protection did not apply to works created by AI. Therefore, copyright laws should protect AI-generated work in order to preserve the spirit of competition in the market. Certain advantages entice the protection of AI-generated works similar to that of human-generated works. Similar to how AI technology has revolutionized the intellectual property sector, consumers will be able to purchase the produced good at a reasonable cost instead of the creator’s exorbitant price. For these kinds of AI-generated works, protection is crucial since, in this scenario, if AI produces cheap goods in big quantities, there is a chance that they could be used by others without giving credit to the AI. Hence it is potentially essential to copyright such creations.
Authorship and Ownership of AI-generated work
When works created by AI meet copyright requirements, authorship becomes crucial. AI systems can now independently create without human input, in contrast to traditional computer-generated works, which presents challenges for authorship attribution. Given AI’s autonomy in making decisions during creation, this calls into question the definition of authorship and AI’s role as an author. Investigating these questions is crucial to figuring out whether AI qualifies as an author and how copyright regulations should change to account for works created by AI.
If AI is thought to be similar to human authors, then the next issue would be joint authorship. It is defined under the Indian Copyright Act, 1957[14]. This clause states that the AI’s contribution is equal to that of other authors and should not be distinguished from them. But if AI generates as a joint author, the other author cannot direct AI’s entire course of action, and as a result, the other author will not have control over the final product.
Given that AI-generated works are capable of being protected by copyright since they satisfy all copyrigtability requirements, the ownership of AI-generated works will inevitably become a question. In this sense, a fresh uncertainty about ‘to whom’ will arise once more. “To whom” in this case refers to the individual or AI that will be in possession of the copyright, and it will be a significant issue.
If the programmer is to be considered as the owner, it can be argued that the AI is programmed in such a way that it follows the program fed into it and hence it can be presumed that it is the creation of the programmer, regarding him as the owner of the work. In such a scenario, question arises when AI is not under the supervision of the programmer or if the works are done after the programmer’s death.
The basic principle of copyright law that the original owner of the work is the creator is broken by the “doctrine of work made for hire.” According to the theory, ownership rights are not given to the person who created or fixed the expression, but rather to the employer or commissioning party who bears the cost of producing the work. The question of whether the person who creates the AI, the owner or creator, should be regarded as the owner of the work emerges if AI cannot be considered an author of the work because it does not meet human requirements. As the AI’s creator invested a great deal of time, money, and expertise in its development, awarding ownership to him will act as a reward. But such copyright does not give recognition to AI.
Attribution of copyrights to the end user can be the next alternative. The end user who makes the needed arrangements to create the work can be regarded as the owner just as one uses MS Word where he would be the owner not the MS Word. However that same rationale is not applicable in the situation where the AI generates the works by employing algorithm and creates on its own. The role of the user is limited to pressing the button and instructing the AI to create the work on its own.
Challenges
After acknowledging that AI works can be granted ownership to the AI, the next concern is when the AI might violate third parties’ rights. Because AI is capable of self-learning, it would be difficult to stop AI from using copyrighted material that is available online if it turned out to be an honest copying and not constitute infringement. The questions that need to be addressed are how to hold AI accountable for its actions.
Right to sue and to be sued is one of the rights offered by the copyright statute to the author. With this regard, how can an AI seek to compensate and prevent others from using its work?
The next issue comes with regard to the protection period. Most of the nations grant copyright protection only for a limited period of time with India granting protection of 60 years following the author’s death under Section 22 of the Indian Copyrights Act 1957. This rule is limited to AI as it is immortal. Hence the copyright protection for Works generated by AI may extend till perpetuity.
The ‘labour theory’ provides that every act of labour and skill performed by a person must be compensated. This has been embraced by every country as a fundamental goal of copyright to reward the creator and encourage them. In case of AI-generated works, the problem arises in rewarding the AI and whether AI is capable of experiencing such reward or encouragement.
Conclusion
The wide range of uses of AI, including self-driving cars, games, and healthcare, present difficulties for copyright regulations that have historically protected works of human creation. As AI-generated works proliferate, concerns regarding their copyright protection surface. In addressing these concerns, the World Intellectual Property Organization (WIPO) emphasized the necessity of legal frameworks to safeguard AI-generated works, which are critical to market competition.
The use of AI in creative processes complicates copyright laws regarding authorship and ownership. Conventional laws acknowledge human creators, but the autonomous nature of AI calls into question this. It is difficult to identify the author of an AI, be it the owner of the AI or the programmer. Possible solutions include acknowledging AI as an author and taking joint authorship into consideration. Ownership rights may transfer to end users, programmers, or work-for-hire. Holding AI responsible for copyright violations, figuring out protection durations, and dealing with the “labour theory” for compensating creators are some of the challenges. The present legislation must be reformed to address the same.
References
- Guadamuz, A. (2017). Artificial intelligence and copyright. WIPO Magazine. https://www.wipo.int/wipo_magazine/en/2017/05/article_0003.html (accessed on 23 June 2024)
- Organization, W. I. (2020). Retrieved from WIPO conversation on intellectual property (IP) and artificial intelligence (AI), second session.: https://www.wipo.int/meetings/en/details.jsp?meeting_id=55309 (accessed on 23 June 2024)
- International Organization for Standardization. What is AI? Retrieved from https://www.iso.org/artificial-intelligence/what-is-ai (accessed on 23 June 2024)
- Narula, G. (2018). Everyday Examples of Artificial Intelligence and Machine Learning. Tech Emergence. https://www.techemergence.com/everyday-examples-of-ai/ (accessed on 23 June 2024)
- IBM. Deep learning. Retrieved from https://www.ibm.com/topics/deep-learning (accessed on 23 June 2024)
- World Trade Organization. (n.d.). Intellectual property: Protection and enforcement. Retrieved from https://www.wto.org/english/tratop_e/trips_e/intel1_e.htm (accessed on 23 June 2024)
- World Intellectual Property Organization. About intellectual property. Retrieved from https://www.wipo.int/about-ip/en/ (accessed on 23 June 2024)
- Jones, J. (2018, October 26). A portrait created by AI just sold for $432,000. The Guardian News. https://www.theguardian.com/artanddesign/shortcuts/2018/oct/26/call-that-art-can-a-computer-be-a-painter (accessed on 23 June 2024)
[1] Wold Intellectual Property Organization. (2020). WIPO conversation on intellectual property (IP) and artificial intelligence (AI), second session. Retrieved from https://www.wipo.int/meetings/en/details.jsp?meeting_id=55309 (accessed on 23 Jun. 2024)
[2] International Organization for Standardization. (n.d.). What is AI? Retrieved from https://www.iso.org/artificial-intelligence/what-is-ai (accessed on 23 Jun. 2024)
[3] Narula, G. (2018). Everyday Examples of Artificial Intelligence and Machine Learning. Tech Emergence. https://www.techemergence.com/everyday-examples-of-ai/ (accessed on 23 June 2024)
[4] IBM. Deep learning. Retrieved from https://www.ibm.com/topics/deep-learning (accessed on 23 June 2024)
[5] World Trade Organization. (n.d.). Intellectual property: Protection and enforcement. Retrieved from https://www.wto.org/english/tratop_e/trips_e/intel1_e.htm (accessed on 23 June 2024)
[6] World Intellectual Property Organization. About intellectual property. Retrieved from https://www.wipo.int/about-ip/en/ (accessed on 23 June 2024)
[7] Copyright Act of India, 1957
[8] Section 2 (1) Trademark Act, 1970
[9] Section 2 (1) (e) Geographical Indications of Goods Act, 1999
[10] Section 22 Design Act, 2000
[11] Jones, J. (2018, October 26). A portrait created by AI just sold for $432,000. The Guardian News. https://www.theguardian.com/artanddesign/shortcuts/2018/oct/26/call-that-art-can-a-computer-be-a-painter (accessed on 23 June 2024)
[12] Naruto v. Slater, 2018 WL 1902414.
[13] Naidoo, M. (2021, August 11). In a world first, South Africa grants patent to an AI system. The Hindu News. https://www.thehindu.com/sci-tech/technology/in-a-world-first-south-africa-grants-patent-to-an-artificial-intelligence-system/article35817497.ece (accessed on 23 June 2024)
[14] Section 2 (z) “work of joint authorship” means a work produced by the collaboration of two or more authors in which the contribution of one author is not distinct from the contribution of the other author or authors;
Author: Keerthana R Reddy, a 3rd Year student at Christ Academy Institute of Law, Bengaluru
