Authorship and Ownership of AI-Generated Works

9th November, 2024 - 6:30 am Categories: AI 0 Comments

Introduction

AI-generated works encompass paintings, poetry, images, sculptures, and writings or any digital content created with the assistance of AI, guided by appropriate prompts from a human. Generative AI models have gradually taken over the creative field. Tools like DALL-E, Gemini, and ChatGPT have bridged human creativity and Machine computation. These tools have gone from performing simple tasks to making original content further blurring the line between humans and machines. Authorship and ownership are key determiners of rights and responsibilities arising out of copyright. They determine who has the moral right and the right to modify, transfer, distribute, licence etc. Generative AI models often consume information to generate content; sometimes the information that it processes could be copyrighted work which could lead to copyright infringement. Who will take the responsibility in this case? The developer of Generative AI? The owner? or the user? A major issue arises here. Mostly all these tools provide in their terms of use that the user owns the content that they generate through these tools. However, if the content generated is not copyrightable in the first place then the terms of use do not do much to change that situation. Copyright is granted only to humans and various traditional Intellectual Property frameworks rely on creativity, originality as well as innovation. In this blog we will understand this concept further by analysing a couple of cases.

Naruto v Slater

It is often regarded as one of the comical cases of Intellectual property law where Naruto an Indonesian crested black Macaque took a selfie with a camera left alone by David Slater, a nature photographer.¹ These pictures became popularly known as the “Monkey Selfies” and later Slater sold the selfies and made the selfie the cover for his book Wildlife Personalities. PETA(People for the Ethical Treatment of Animals) later claimed that the monkey was the owner of the photos and hence should be given the copyright. PETA utilised the doctrine of next friend status to argue on behalf of the monkey. Here the owner of the camera was David Slater and the author or creator of the photo was the monkey. The US Court of Appeals for the Ninth Circuit established

¹ 2018 SCC OnLine US CA 9C 94

that an animal, not being a human entity could not hold a copyright or any intellectual property rights. It was also held that the doctrine of next friend status by PETA was insufficient.

In both scenarios, the “author” is a non-human entity that challenges the traditional requirement for copyright holders to be human, demonstrating creativity and imagination. Unlike Naruto who accidentally took the photograph, Generative AI processes information, analyses data using a variety of algorithms according to various prompts, and provides an output which is resourceful to the user. Generative AI is programmed to create. Even though it does not possess an intent, it functions in response to human intervention and has a link to human purpose as the user’s design prompts to receive desired outputs. Generative AI tools have a direct purpose and structured human intervention which sets it apart from accidental acts by an animal. AI uses vast databases, machine learning, and various algorithms programmed to imitate human creativity and even alter its output according to the feedback received to meet the requirements of the user.

Thaler v Comptroller General of Patents, Designs, and Trademarks

This case explores the question of whether an inventor status could be given to an AI tool under UK patent law. Dr. Stephen Thaler developed DABUS(Device for the Autonomous Bootstrapping of Unified Sentience)² which he claimed was autonomous and independently invented a flashing light beacon to attract attention during an emergency and a new kind of food or beverage container. Thaler claimed that he was the rightful owner of DABUS. He sought a patent which granted the AI tool named DABUS an inventor status. The United Kingdom Intellectual Property Office denied his application stating that inventorship can only be given to a natural person. In the UK Patents Act 1977 ‘inventor’ in relation to an invention means the actual deviser of the invention which the law interprets to be a human.

Later South African companies and Intellectual Property Commissions became the first country to grant a patent to DABUS and give it inventor status but the ruling was eventually revoked. A similar barrier exists in both copyright law as well as patent law for the limitation of non-human contributors. This case highlights the need for human intent and inputs for AI tools to function. Thaler’s case reveals the rigidity in the patent law for requiring a natural person to be an inventor.

The copyright system rewards and protects creative human intellect. It provides the creator with financial gain when others find the creator’s work resourceful and use the work. It also protects the creator by penalising those who try to circumvent the system. If AI-generated work is not given copyright only because of the lack of a human element then they could be exploited by anyone. For example, it is like writing on sand, the user can call it his but without legislation to

protect it, it could be replicated by anyone. This is one of the reasons why there is a chilling effect among those who are willing to invest in a generative AI tool. Commonwealth countries like New Zealand and the United Kingdom, have actively upgraded their copyright legislations to include technology and AI in creative works.³ The EU has acknowledged the requirement to discuss the role of AI in generating intellectual property but has stayed silent in the aspect of ownership and authorship to AI. Hybrid ownership of copyright could also be a possibility where both the developer as well as the user gets to enjoy the benefits of copyright. As the world catches up with innovation, the resulting legal ambiguity impacts all sides of the AI equation – developers, content creators and copyright owners.⁴ Technology and AI are advancing at a rapid pace with each passing day. The lawmakers should take a step and close the lacuna that has been expanding. They should create an adaptable and technologically aware framework to protect both human and AI driven creativity.

Author: Manjumol R S, in case of any queries please contact/write back to us via email to chhavi@khuranaandkhurana.com or at  Khurana & Khurana, Advocates and IP Attorney.

³ Corrs Chambers Westgarth, “Artificial intelligence and copyright: ownership issues in the digital age”, Lexology’s Website, https://www.lexology.com/library/detail.aspx?g=849627a6-c428-4e45-

a386-c6e49d98b446,

⁴ Will copyright law enable or inhibit generative AI https://www.weforum.org/agenda/2024/01/cracking-the-code-generative-ai-and-intellectual-property/