Historical Context of Copyright Law
- Copyright law originated in 1710 due to the invention of the printing press.
- Its aim: protect publishers’ rights, encourage learning, and secure economic interests.
- Over time, it has adapted to photocopying, recording devices, and the Internet.
- Each technological shift brings debates on how copyright should respond.
Relevance : GS 2(Governance) ,GS 3(Invention ,Technology)
Generative AI: A New Challenge
- Current concern: generative AI trains on copyrighted material without permission.
- This shifts focus from “copying works” to “training on works.”
- Earlier legal concerns involved reproducing copies; now it involves usage in training datasets.
Global Legal Crossroads
- Generative AI companies (like OpenAI) use internet scraping to collect both copyrighted and non-copyrighted content.
- Lawsuits have emerged globally:
- India: Federation of Indian Publishers & ANI sued OpenAI in Delhi HC.
- USA: Claims countered with “fair use in education” exceptions.
- OpenAI introduced an opt-out mechanism, but it only applies to future training, not past.
India’s Unique Legal Landscape
- India follows an enumerated exceptions model under its Copyright Act.
- Unlike the U.S. “fair use” doctrine, India lists specific exceptions—limited scope.
- Educational use is confined to classroom settings—favouring right-holders in disputes.
- Indian courts may face jurisdictional challenges, but the core issue remains unresolved.
Key Judicial Considerations
- Amicus curiae (Dr. Arul George Scaria) suggestions:
- Assess feasibility of “unlearning” content already used in training.
- Balance AI development with access to legitimate information.
- Address false attribution issues in AI responses.
Concerns about Access and Equity
- Over-restriction may hurt access to books and knowledge—undermining copyright’s original intent.
- Newer, smaller AI players could suffer due to lack of access to high-quality training data.
- Courts must ensure a level playing field between dominant and emerging AI platforms.
Foundational Copyright Principles as a Guide
- Copyright protects expression of ideas, not the idea/information itself.
- If AI uses information (not expression), it’s not necessarily infringement.
- Law should distinguish between:
- Learning from content (permissible)
- Copying protected expressions (infringement)
Philosophical and Practical Implications
- All creativity—human or AI—is based on learning from the past.
- Creating a legal divide between human and machine learning may be counterproductive.
- Law must evolve but not at the cost of stifling creativity and future innovation.
