Generative AI & Issues of Copyright

Syllabus: GS3/Science & Technology

Context

• Generative AI does not merely reproduce copyrighted works but trains on them, raising concerns about unauthorized usage, unlike previous technologies.

About the Generative AI

• It refers to AI systems capable of creating new data, whether it’s text, images, or code. It is driven by advancements in Large Language Models (LLMs) that have the capability to generate new data, whether it’s text, images, or code.
• Generative AI models are trained on massive datasets, often scraped from the open internet. These datasets frequently include copyrighted material, sometimes without the explicit consent of the copyright holders.
  • GenAI tools are now being used in mainstream journalism, advertising, entertainment, and education.
• It has raised ethical and legal concerns over whether the use of such data in training AI constitutes fair use or a breach of copyright law.

Evolution of Copyright Law

• Statute of Anne (1710): It was the world’s first copyright law enacted in England, that introduced the concept of the author of a work being the owner of its copyright, and laid out fixed terms of protection.
  • It established fixed terms of protection and required registration at Stationers’ Hall.
  • It also led to the Copyright Act of 1790 in the United States.
• Berne Convention (1886): It created international copyright standards, ensuring mutual recognition of copyrights across nations, and eliminated the need for separate registrations in different countries.
  • It remains in force to this day, and continues to provide the basis for international copyright law.
  • India is part of the Berne Convention.
• Copyright Registration Systems: Registration systems vary across countries, while the Berne Convention protects unpublished works.
  • Some nations offer optional registration, while others rely on automatic copyright protection.

Copyright Issues in Generative AI

• Unauthorized Use of Copyrighted Content: AI companies use web scraping methods to train their LLMs on a vast array of data, including both public and copyrighted content.
• Fair Use Debate: In the USA, OpenAI faces similar lawsuits, where it has invoked ‘fair use’ and ‘fair learning in education’ as defenses under American copyright law.
  • However, OpenAI has introduced an opt-out mechanism for future training, but it does not address past usage.
• Legal Challenges in India: The Federation of Indian Publishers and Asian News International have filed lawsuits against OpenAI in the Delhi High Court, alleging unauthorized use of their works.
• Music Industry Concerns: Bollywood music labels have joined copyright lawsuits against AI platforms, citing unauthorized use of sound recordings for AI training.

Global Legal Ambiguity

• United States: It has clarified that purely AI-generated works are not eligible for copyright protection.
  • It has prompted creators to include ‘substantial human authorship’ in AI-assisted works to ensure copyrightability.
• Japan: It has explicitly stated that AI training using copyrighted data does not infringe copyright as long as it is non-consumptive and for machine learning purposes.

Key Legislations in India

• Copyright Act, 1957: The Copyright Act of 1957 does not directly address AI-generated works.
  • Section 2(d) of the Act defines an ‘author’ in human terms, making it difficult to attribute authorship to a non-human agent.
• Panel Review on AI & Copyright: The Indian government has set up a panel to assess whether existing copyright laws are sufficient to regulate AI-related disputes.
• No Separate IPRs for AI Content: The government has clarified that there is no proposal to create separate Intellectual Property Rights (IPRs) for AI-generated content, relying on existing copyright and patent laws.

Legal Complexities in India

• India’s copyright framework is significantly different from the US model, as:
  • India follows an enumerated exceptions approach, not the flexible US ‘fair use’ test.
  • Educational exceptions in India are narrowly defined, confined to classroom use.
• It was urged by the Delhi High Court to consider whether it is technically feasible for AI to ‘unlearn’ previously absorbed copyrighted data.
• It limits maneuverability for AI developers and may favor right-holders in litigation.

Way Forward: Toward Ethical AI Creation

• Regulatory frameworks need to evolve and ensure that original human creators are respected, credited, and, where appropriate, compensated.
• Balancing Innovation and Protection: There is a need for a level playing field to ensure fair competition in the AI space.
• Proposals gaining traction include:
  • AI transparency laws, requiring companies to disclose datasets used in training.
  • Opt-out registries for creators who do not wish to have their content used.
  • Fair licensing schemes that allow AI companies to use copyrighted data with remuneration.

Conclusion

• Copyright law stands at a pivotal moment. Generative AI challenges traditional legal interpretations, but its regulation must not hinder creativity and access to knowledge.
• By reaffirming the core principles of copyright and ensuring fair treatment of all players in the AI ecosystem, the law can continue to serve its dual purpose—protecting creators while promoting learning and innovation.

Source: TH