Copyright law protects original works, including computer software, under India's Copyright Act of 1957. While AI-generated works raise questions about authorship, the Act doesn’t explicitly address them. As AI advances, copyright law may need reform to accommodate these changes.
Evolution, Purpose, and Rationale of Copyright Law
Historically, copyright law has been primarily associated with artistic and literary works, but in modern times, it has evolved to provide protection for a broader range of creations, including computer software. The primary purpose of copyright law is to safeguard the interests of creators by protecting both their moral and economic rights. As noted by the distinguished legal scholar Miller, the rationale for copyright is to incentivize creators to produce original works, thereby enriching society and generating revenue from these creations. Copyright law thus plays a crucial role in encouraging innovation and ensuring that creators can reap the benefits of their intellectual endeavours.
Under the Copyright Act of 1957 in India, copyright protection is granted to original works that are fixed in any tangible medium of expression. This includes both the protection of the creator’s exclusive rights and the prevention of unauthorized use of their work. The Act allows the copyright owner exclusive rights to reproduce, create derivative works, and distribute copies of their work. Importantly, there is no requirement for registration or any formal steps after a work has been fixed in a tangible medium to enjoy copyright protection. However, recording the copyright of a work is essential for initiating legal action in the case of infringement.
Copyright Protection for Computer-Generated Works
In terms of computer software, which may not necessarily have a methodological effect, it is protected under the copyright law if it meets the criteria of originality. The software must be the result of significant effort and skill to establish its uniqueness. For instance, a program that merely generates multiplication tables would not meet the required level of effort to warrant copyright protection. Additionally, for software to be protected under the Copyright Act, it must be either created in India or, if created outside India, the author must be an Indian citizen or domiciled in India at the time of creation or publication.
The protection of computer programs as literary works hinges on their originality, meaning that the work must not be copied from existing sources but should instead reflect the creator’s independent contribution. This principle is applicable to software, where the creator is the individual responsible for its generation. While computer programs are protected under copyright law, simple programs that require minimal skill, such as those generating basic algorithms, may not qualify for protection. The protection process generally follows several stages:
- Generating Conceptual Preliminary Material: This step involves brainstorming and drafting the initial ideas for the software.
- Generating the Source Code: This involves writing the program in a specific programming language.
- Generating the Object Code: Here, the source code is compiled into a machine-readable format.
- Creating User Manuals: A user manual or documentation for the program is produced to guide end-users in using the software.
Machine Learning, AI, and Copyright Issues
The development of machine learning (ML) and artificial intelligence (AI) technologies introduces complex challenges in the realm of copyright protection. AI and ML are increasingly used in diverse fields such as robotics, cloud computing, healthcare, and entertainment, with applications ranging from facial recognition to content creation. As these technologies advance, questions regarding the copyrightability of works generated by AI systems have emerged.
The Copyright Act of 1957 in India provides copyright protection to both the object code and the source code of software, including AI applications, as part of its literary works definition. However, the issue becomes more complicated when considering works generated by AI without significant human involvement. In this context, AI-generated works raise profound questions about authorship and the applicability of copyright law.
Recent legal cases, such as the Telstra v. Phone Directories[1] case, have explored whether works created by computer programs or AI systems can be considered original literary works. The case focused on determining the authorship and originality of computer-generated content. The increasing sophistication of AI technology, particularly generative AI, has further complicated the issue of copyright protection. Works produced by AI, such as those generated by an algorithm based on prompts or input data, are often seen as outputs of machine processes rather than products of human creativity.
In the European Union, courts have ruled that AI-generated works cannot be considered original works because they lack the intellectual creation typically associated with human authorship. The Infopaq[2] case, in particular, emphasized that copyright applies only to original creations that reflect the intellectual effort of the author. AI-generated works are not protected under current copyright laws, as they do not meet the requirement of human authorship.
However, some jurisdictions, such as China, have taken a more nuanced approach. In certain cases, AI-generated works may qualify for copyright protection if sufficient human oversight is involved in the creation process. In India, the Copyright Act of 1957 does not explicitly address AI-generated works, but it does define "computer-generated works" under Section 2(o)[3]. This definition includes computer programs, databases, and tables, but it does not clearly extend to works generated entirely by AI systems.
The Indian Copyright Act allows for the person who causes a computer-generated work to be created to be considered the author. This provision may apply to AI-generated works, where the entity or individual responsible for designing and training the AI system might be seen as the creator. However, the ambiguity surrounding AI authorship remains a significant challenge in determining who holds copyright in such works.
Legal Considerations for AI-Generated Works
As AI continues to develop and generate content, there is a growing need for copyright laws to adapt to the realities of synthetic creativity. Current laws, both in India and internationally, struggle to address the unique characteristics of AI-generated content. For example, while AI outcomes are based on complex algorithms that predict the likelihood of certain responses or results, they do not involve the type of intellectual creation associated with human authorship. Therefore, AI-generated works may be better understood as outputs of machine processes rather than creative works deserving of copyright protection.
In some cases, courts have ruled that AI-generated works, such as those created using machine learning or other AI technologies, do not qualify for copyright because they lack the human intellectual input required to establish authorship. This raises the question of whether the creators of the AI systems themselves should be granted copyright over the works generated by the system. Some legal scholars, argue that while AI may be used as a tool for generating new content, the ownership of the resulting work should likely rest with the human creators of the AI systems, rather than the machine itself.
Conclusion
The intersection of AI, machine learning, and copyright law is a rapidly evolving area of legal practice. The current copyright framework in India and globally was not designed to address the complexities introduced by AI-generated works, leading to ongoing debates over authorship, originality, and ownership. While the Copyright Act of 1957 provides protection for computer programs as literary works, it does not explicitly account for works generated solely by AI. As AI technology continues to advance and create content, there will likely be increased pressure to reform copyright laws to address these emerging challenges and ensure that creators, whether human or machine-assisted, can protect their intellectual property in an increasingly digital world.
[1] Telstra Corporation Limited v Phone Directories Company Pty Ltd, [2010] FCA 44.
[2] Infopaq International A/S v Danske Dagblades Forening, (C-5/08).
[3] The Copyright Act, 1957, s. 2(o).
