AI AND IPR: WHO GETS OWNERSHIP OVER AI- GENERATED WORK

ABSTRACT

The rise of Artificial Intelligence (AI) has challenged traditional frameworks of intellectual property rights (IPR), especially in determining authorship and ownership of AI-generated works. This article talks about the growth of AI, its relevance in IPR Laws and the legal framework relating to the ownership of property rights over AI generated content in four major countries. The article mainly focuses on the question that ‘ who shall possess the ownership rights of a work which has been created/generated solely by AI and without any human intervention?’.

INTRODUCTION

Artificial Intelligence has become an indispensable part of our life. From students to professionals, everyone has started relying on AI for their chores which is not a bad thing since, AI helps reduce the workload and increase one’s efficiency. AI has now entered all those areas where humans are present and can perform the tasks even better than us. AI is integrating as well as replacing humans at the same time which raises the question- In case of a creation by AI, who shall be given the ownership of such creation? Is AI capable of being entitled to  ownership rights too? Across nations, IPR Laws have been framed keeping in mind that the creations have been made by humans and thus they recognise only humans as owners of IPR. However, in today’s time AI has developed so much that it is capable of composing music, writing news articles, generating artwork, and even invent novel technological solutions. These developments raise an essential legal question: who owns the rights to these AI-generated works? This article shall deal with the various aspects relating to this question.

ARTIFICIAL INTELLIGENCE

The term ‘Artificial Intelligence’ was first used in a 1956 conference by Mr. John McCarthy, a computer scientist. He claimed that it was the idea of a computer analysing information and behaving on it in a way that is similar to how an individual (with an intellect) would react to the same input^[1].

Artificial intelligence (AI) is the theory and application of computer systems being able to accomplish tasks traditionally requiring human intelligence, like the recognition of speech, decision-making, and pattern recognition. AI is a broad term that covers many different types of technologies, including deep learning, machine learning, and natural language processing (NLP).

Some of the commonly used forms of AI are:

• ChatGPT: Uses large language models (LLMs) to generate text in response to questions or comments posed to it.
• Google Translate: Uses deep learning algorithms to translate text from one language to another.
• Netflix: Uses machine learning algorithms to create personalized recommendation engines for users based on their previous viewing history.
• Tesla: Uses computer vision to power self-driving features on their cars.

Turing Test: There were doubts regarding whether the output generated by AI is a result of the algorithm or through the machine’s own intelligence. To clear out the air, Turing Test was introduced wherein a human judge engages in a text- based conversation with a human and a machine. Through the responses of the machine and human, the judge has to determine whether the response had been given by a human or machine. The machine has to fool the judge into thinking that the response it generates is that of a human. The machine passes the turing test if the judge isn’t able to decipher the response of the machine.

If the machine would pass the test, it could be ascertained that the output of AI is through its own intelligence. However, the test faced severe criticism due to its limited scope, human bias, for portraying a non- realistic version of AI.

AI AND IPR

Intellectual Property Rights serve as an incentive to people for creating, innovating new products, designs, music, etc. However, these laws have been formulated keeping in mind that such tasks can only be performed by humans. But as the era of AI begins, so does the ambiguity comes into the picture.

The involvement of technology has become vital for humans. AI contributes to the development of a product and also develops products completely on its own. The key characteristics of AI generated works are:

• Autonomy in content generation
• Absence of human emotional or intellectual input in the final product
• Reliance on existing datasets, which may themselves be protected under IPR

With such a huge role being played by AI in the IP field, there is still no clear cut answer for ‘who possesses the IP rights for a substance created by their invention’

GLOBAL STATUS

The article shall now delve into the legal frameworks relating to the same in various countries.

INDIA

Despite being the second most populated country and among the top developing countries, India is yet to come up with laws for AI generated content. The ambit of AI is highly unclear. The Indian Copyright Act, 1957 however pours some light over the issue.

Section 2(d) of Indian Copyright Act, 1957 defines ‘author’. Section 2(d)(vi) recognises a person who causes the computer-generated work to be created as the author of the work. The act does not give any rights to the software/ AI. The problems in this provision was brought forward in the RAGHAV Case(2020)  wherein the AI System named “RAGHAV’ sought to get copyright registration from the Indian Copyright Office for its artwork called ‘SURYAST’. The request was initially denied but later accepted when a natural person was named as a co-author alongside ‘RAGHAV.’ Later on, a withdrawal notice was issued, seeking clarification on the legal status of ‘RAGHAV,’ highlighting the ambiguity surrounding AI’s qualification as an artist under the Copyright Act.

The Patent Act, 1970 and Design Act, 2000 do not provide any express provision regarding AI generated work. Section 2(p) and 2(t) of the Patent Act, 1970 mentions the terms “patentee” and “person interested” which rules out the possibility of including AI in the said provisions. The act does not recognise the creations by computer programs, business methods or mathematical formulae as patentable inventions.

USA

The copyright laws are governed by the Copyright Act, 1976. The act recognises only humans as a creator and does not recognise the work produced by a machine or mere mechanical process that operates randomly or automatically without any creative input or intervention from a human author. AI is considered as a tool for producing the work rather than an author itself.

The US Supreme Court in Mohamad v Palestinian Auth^[2] laid down in its judgement that Section 115(b)(2) of the act which talks about inventor’s oath or declaration has mentioned ‘himself’ and ‘herself’ rather than ‘itself’. Thus, machines cannot be considered as authors and be granted copyright rights. A similar stance was taken by the court in Thaler v. Perlmutter (2023)^[3], where the court held that copyright law protects only works of human authorship, reaffirming the denial of copyright registration to a piece of art generated by an AI system known as “Creativity Machine.” The court emphasized that human creativity is a constitutional and statutory prerequisite for copyright protection in the United States.

CHINA

China has adopted a comparatively progressive and pragmatic course regarding the legal status of AI-created works under its intellectual property law. Although its 2020 amended Copyright Law formally demands human authorship to be protected, Chinese courts have been flexible in interpreting this mandate. An exemplar case in point is that of Feilin v. Baidu (2020), whereby a Beijing court extended copyright protection to an article composed by Tencent’s AI application, Dreamwriter, on the basis that the work evidenced sufficient originality and that human developers had an overseer’s function. This represented a significant progression toward the identification of AI-powered creations, particularly where there exists evident human presence in the composition. Further, policy papers such as the White Paper on Artificial Intelligence and Intellectual Property by China National Intellectual Property Administration (CNIPA) 2020 support adjusting legal standards in line with changing technology, and this shows the desire of China to modify its IP laws for the purpose of supporting innovation and maintaining the basic principles of originality and authorship.

JAPAN

According to the Japanese Copyright Act, copyright is only afforded to works embodying human intellectual creativity, which means works made independently by artificial intelligence (AI) systems are excluded. If an AI produces content without substantial human input—like writing music, articles, or images—such work is not subject to copyright protection under existing Japanese law. This policy aligns with Japan’s focus on the human mind as being at the heart of the creative process.

However, Japan has taken forward-looking steps to promote AI development, especially in data use. Article 30-4 of the Copyright Act states an invaluable exception by enabling the utilization of copyrighted material for analysis of information—including machine learning and AI training—without any infringement of copyright, if only the use is not to appreciate the original form. The said provision has facilitated researchers, developers, and tech firms to mine big data lawfully, which is the quintessential necessity in training advanced AI systems.

In patent law as well, Japan maintains the need for human inventorship. AI systems cannot be credited as inventors; however, patents can be awarded on AI-enabled inventions if they comprise substantial human involvement and a human inventor is included on the filing. This provides assurance that even though AI may enable innovation, legal rights and obligations are reserved for humans or legal persons like corporations.

In spite of these limitations, Japan is proactively looking into future reforms to deal with the increasing use of AI in content creation and innovation. The Japan Patent Office (JPO) and the Agency for Cultural Affairs have carried out public consultations and published policy papers examining the implications of AI on IPR. These consultations suggest a future-oriented approach, with the potential to create new legal categories or sui generis protections specific to AI-generated works.

CONCLUSION

AI has now achieved heights which no one could have imagined 10 years back. AI is powerful enough to replicate human actions and perform all those task which a human can. The issue which has been discussed in the article shall be dealt with utmost urgency. Granting IP rights to AI however poses some challenges like dominant position being exploited by companies who have a strong AI system and database. Also, the information used by AI might have been protected and through its use, AI is committing a severe infringement. Also, there is no uniform law across the nation which would again create ambiguity in this globalised era. Thus, it is important that policymakers, technologists, and legal scholars globally collaborate to ensure that the legal architecture evolves in tandem with technological progress.

REFERENCE(S)

1. Tian Lu, Feilin v. Baidu: Beijing Internet Court Tackles Protection of AI/Software-Generated Work and Holds That Copyright Only Vests in Works by Human Authors, The IPKat (Nov. 3, 2019), https://ipkitten.blogspot.com/2019/11/feilin-v-baidu-beijing-internet-court.html.
2. U.S. Copyright Office Guidelines, 2022
3. Indian Copyright Act, 1957
4. White Paper on Intellectual Property and Artificial Intelligence, China (2020)
5. Japanese Copyright Act

^[1]  T.H. Cormen “Algorithms Unlocked” (The MIT Press Cambridge, Massachusetts London, 2013).

^[2] Mohamad v. Palestinian Auth., 566 U.S. 449 (2012)

^[3] Thaler v. Perlmutter, No. 22-cv-1564 (BAH), 2023 WL 5333236 (D.D.C. Aug. 18, 2023)