INTRODUCTION
ChatGPT and similar generative AI models have smashed down on the creative industries like the fist of God, by enabling the rapid production of text, images, and music. While these technologies have paved the way for new forms of artistic expression, they also bring with them a whole slew of copyright challenges that require rethinking traditional intellectual property laws. In India, copyright is governed by the Copyright Act of 1957, which protects original works of literature, art, and drama, inter alia. This protection is granted based on originality, a concept that has been clarified through Indian jurisprudence to require a "modicum of creativity" rather than blunt elbow grease. In Eastern Book Company v. D.B. Modak[i], the Supreme Court of India emphasized that originality involves a degree of skill, judgment, and labour, distinguishing it from purely mechanical effort and the definition of "authorship" under Indian law remains tied to human creators. Well, the part about ‘human creators’ was not spelt out in clear letters but as the jurisprudence has evolved, it has become clearer. This principle was famously tested in the Macau selfie case, wherein a (now) pretty famous monkey accidentally took a photograph using a camera set up by a photographer. The courts ruled that the monkey could not be the author of the photograph, as copyright law requires authorship to be attributed to a natural person. This set forth a clear principle, that no entity be it an animal or artificial intelligence, can claim copyright, regardless of its demonstrated capabilities or the resemblance of its output to human creations.
CHALLENGES WITH GENERATIVE AI
The advent of ChatGPT and other generative AI models has disrupted this framework by introducing new forms of creativity that lack clear human attribution. AI models are trained on vast datasets, often scraped from the internet, which can include copyrighted material. This casts doubt about whether AI-generated outputs qualify as "original" and who, if at all anyone, owns the copyright to such works. These issues are further compounded by the fact that AI-generated content often mimics patterns and styles learned from preexisting works, challenging the concept of originality as a purely human endeavour. The use of copyrighted material in training datasets without explicit consent, for example, in Getty Images v. Stability AI[ii], Getty alleged that Stability AI had used over 12 million copyrighted images from its database to train its models without authorization. This case underscores the tensions between fair use provisions and the expansive data requirements of generative AI systems. Such practices have drawn criticism from artists, who argue that their work is being exploited to train AI systems without consent. Polish artist Greg Rutkowski, whose name was used over 93,000 times to prompt AI-generated art, serves as a prominent example of this concern. Adding to these challenges is the ethical dilemma posed by AI's ability to generate content that competes with human creators. ChatGPT can produce text, images, and music that rival human work in quality, raising fears of job displacement in creative industries.
LEGAL AND ETHICAL IMPLICATIONS
The levee broke during the 2023 SAG-AFTRA strike in the United States, where actors and writers protested the unregulated use of AI in entertainment. Among their concerns was the potential for AI-generated performances and scripts to undermine the livelihood of human creators. This strike highlighted the urgent need for stronger regulations to ensure fair treatment and equitable compensation in industries affected by generative AI. Current laws in India are unclear about the ability of an AI to hold a copyright, and the matter is, as it stands, a grey area. The recent controversy surrounding RAGHAV AI, where the Indian Copyright Office initially granted and later withdrew copyright protection for the AI-generated artwork ‘Suryast,’ highlights the same. Despite these challenges, current laws in India and many other countries do not recognize AI as a legal entity capable of holding copyrights. However, recent Indian cases highlight the emerging complexities of AI's intersection with existing intellectual property frameworks. For instance, in Arijit Singh vs. Codible Ventures LLP and Ors., the Bombay High Court addressed the misuse of the plaintiff's personality rights. The case involved AI tools that converted the plaintiff’s voice into imitative outputs without authorization, violating his rights to control his likeness, name, and voice. The court held that such unauthorized use of AI to manipulate a celebrity’s persona for commercial gain infringes upon their personality rights, underscoring the importance of protecting human identity and creativity against technological misuse. The Court granted interim injunction to prevent further harm and a suit on the matter was decreed, yet a clear stance on the same cannot be inferred from a mere interim order and it remains to be seen how the case evolves. This patent ambiguity warrants clearer guidelines for defining the rights of data contributors and AI developers, establishing consent protocols for using copyrighted material in training datasets, and determining the scope of fair use in generative AI contexts, and until said guidelines can be formulated. Ethical considerations must also be prioritized, ensuring that AI technologies are designed to respect the rights of creators while fostering innovation.
CONCLUSION
Beyond copyright, financial losses arising from unauthorized use of content by generative AI remain an important concern. For creators, the economic implications include diminished market opportunities, disrupted business models, and potential reputational harm. In the ongoing case of ANI Media v. OpenAI in the Delhi High Court, ANI has alleged that OpenAI used its news content without permission to train its language models, particularly ChatGPT, and further accused the it of replicating ANI’s content verbatim while falsely attributing fabricated news to the agency.
For users of generative AI, the risks include liability for copyright infringement, which could result in damages and compensation claims. For example, under Indian law, statutory punishment for copyright infringement can extend to imprisonment and fines. Generative AI will continue to reshape the creative landscape, but the challenge that it poses must be met by striking a balance between technological progress and the protection of human creativity.
[i] 2008 AIR SCW 49
[ii] [2023] EWHC 3090 (Ch)
Disclaimer:
This post is for informational purposes only and does not constitute legal advice. The contents are based on general legal principles and should not be construed as specific advice for any individual or entity. Readers are advised to seek professional legal counsel tailored to their particular circumstances before taking any action based on the information provided.
The sharing of this post does not create an attorney-client relationship between the authors, the firm, and the readers. While every effort is made to ensure the accuracy of the information at the time of publication, laws and regulations are subject to change, and no liability is accepted for any errors or omissions.
For further assistance or professional advice, please contact Sansa Legal directly.
Comments