As the exponentially growing tech corridors of Mumbai, Delhi, and especially the Silicon Valley of India – Bengaluru, are relaunching the content industry through Artificial Intelligence (AI), ranging from art and music to critical software codes, they are faced with the challenge of navigating ownership and originality in terms of issues of copyright. The intersection of copyright and AI-generated work has triggered a global policy conundrum: firstly the right pertaining to the input data used during the training process of the AI platform; secondly the protection of output and due attribution in the work created by the AI platforms.

The critical issue is that the content created by AI is not original in nature but derived from various sources of original information by a computer-generated code. AI platforms not attributing the source of information create the ambiguity between plagiarism and reproduction of part of the original content.

AI in the creative industry has sparked worldwide excitement and controversy. According to CB Insights, a data company that collects and collates data for business, says globally investments in generative AI grew fivefold from $4.3 billion in 2022 to $21.8 billion in 2023. India is fast catching up to the global trend aspiring to be the global hub for AI with more than 3,000 AI startups by 2023 with projected revenue of $17 billion by 2027.

For the growth to sustain India needs to overhaul its copyright laws of 1957 to address the challenges posed by content generated by AI. Broadly speaking, copyright is a legal protection and exclusive economic and moral rights given to the creator for their original tangible form of intellectual expression. Specifically, it is crucial to understand the defence of fair use and fair dealing while discussing the copyright issues of generative AI. Fair dealing in India is enshrined in Section 52 of the Copyright Act, 1957 which provides a list of scenarios like research, judicial proceedings, reproduction for educational purposes, and so on as exceptions to copyright protection. Globally the United States has one of the most liberal fair-use doctrine codified in Section 107 of the Copyright Act which considers the purpose, nature, amount, and effect of the use of the copyrighted work upon potential market. In comparison to the US, the European Union provided fewer exceptions hence less flexibility with fair usage of the copyright work.

The Copyright Act of India 1957 which was updated in 2021 has missed out on taking note of the potential challenges that AI could bring in. This is specifically with regard to revenue leakages and violations over content created and owned by individuals and business houses. Even after twelve years, in February 2024, India reiterated the current provisions in the Copyright laws, defending its stand against any new changes for addressing challenges posed by AI. Earlier, in 2018, the National Strategy for Artificial Intelligence prepared by NITI Aayog, the country's official think tank, as well as the draft National Strategy on Blockchain by the Ministry of Electronics and Information Technology (MeitY) did not delve into copyright issues triggered by AI platforms.

The European Union has been the torchbearer for the world and is charting the middle path with its AI Act 2024. It takes a risk management approach by implementing regulations according to the risks involved. This balance between innovation and ethical considerations is seen as a model by several other countries.

The United States still is seen to take a conservative viewpoint. In 2022 the Copyright Office dismissed copyright protection for an AI generated image and making copyright limited to work created by humans only. This potentially restricts innovation in the creative fields of AI in the US. On the other hand, China has a more progressive approach in this regard. As per the 2022 Guidelines from China's National Copyright Administration copyright protection is granted to the AI-generated work, although the nature of authorship and creativity is disputed and is under question.

These different approaches highlight the urgency of exploring global consensus and the complexity of the challenge.

Is there a way forward to resolve this policy ambiguity? Absolutely!

The MeitY’s March 15, 2024 advisory is an example of the way forward in AI regulation in India. It mandates due diligence compliance for the use and deployment of AI tools by intermediary platforms. Still, there is a huge gap as it was just an advisory, not a law passed by the parliament.

India should derive its learning from global examples, and customise it to domestic issues. The outcome could be a concise and comprehensive AI copyright framework, which could be a broader update or extension of the current Copyright Act. There should be a detailed spectrum of rights and protection-based approaches on the share of human creative input, as a significant human intervention should enjoy greater rights in comparison to works with minimal human contribution. Such a framework should encourage innovation and protect creative rights, ensure ease-of-doing-business. That would help India diligently tackle the copyright challenge and become a leader in AI Governance.

Eshita is a lawyer and tech policy analyst, currently doing Masters at The Fletcher School, Tufts University and cross-registered at Harvard University. She is based in Boston, MA, USA.