Relevance: GS-3 (Science & Technology – AI Governance, Digital Economy), GS-2 (IPR, Regulation)
Source: The Hindu; DPIIT Working Paper

News

The Government of India is planning major amendments to the Copyright Act, 1957 to respond to new challenges created by Artificial Intelligence (AI) and Large Language Models (LLMs). 

A DPIIT working paper proposes a blanket licensing framework allowing AI systems to legally access online content while ensuring fair compensation to publishers and creators. This marks India’s first comprehensive attempt to regulate AI training data.

Need for Changes

AI models such as ChatGPT and Gemini train on massive quantities of online content—news, articles, images, books—without direct permission from creators.

This creates regulatory gaps:

• Copyright uncertainty: Whether AI training counts as copyright infringement.
  
• Economic concerns: Publishers lose control over their content used to create AI outputs.
  
• Legal disputes: Courts in the US and Europe are already hearing lawsuits on AI content misuse.
  
• Absence of Indian rules: The current Copyright Act does not define “AI training,” “data scraping,” or “AI-generated works.”
  

A modern digital economy demands clarity, transparency, and fair value-sharing.

Main Proposals for Change (DPIIT)

1. Blanket Licensing Framework

• AI developers can train models using all publicly available online content.
  
• In return, they must pay royalties.
  
• A copyright society will collect and distribute payments—similar to music and broadcasting royalty systems.
  

2. New Rules for AI-generated Works

A second report will explore:

• Whether AI-generated content should enjoy copyright protection.
  
• Who is the author—AI developer, user, or neither.
  

3. Amendments to the Copyright Act, 1957

• Definitions for AI training, datasets, and automated reproduction.
  
• A structured dispute-resolution mechanism for AI-related claims.
  
• Stronger protections for publishers to track use of their content.

Possible Challenges in Implementation

• Royalty calculation complexities: Billions of data points make valuation difficult.
  
• Transparency of AI datasets: Firms may not disclose training sources.
  
• Opt-out ambiguity: Publishers want the right to restrict scraping.
  
• Innovation impact: Over-regulation may disadvantage Indian startups relative to global competitors.
  
• Enforcement difficulties: Tracking misuse in generative AI outputs is technically challenging.
  

Way Ahead

• Transparent dataset registries and disclosure norms.
  
• Clear “opt-out” provisions for sensitive or premium content.
  
• Joint industry–publisher licensing models.
  
• Harmonisation with global AI standards (EU AI Act, OECD AI Principles).
  
• Promote Responsible AI that balances innovation, creators’ rights, and public interest.
  

Copyright reform must protect creators without slowing India’s digital and AI innovation ecosystem.

Q. “Evaluate India’s proposed copyright reforms for AI. How can policymakers balance innovation with the rights of creators?”

Hints
Intro: India’s proposed copyright reforms seek to regulate AI training while protecting creators.

Body: Blanket licensing enables innovation but raises consent, liability, and enforcement concerns.

Value Addition: WIPO debates, EU AI Act, opt-out model, copyright society for royalties.

Conclusion: Balanced rules must ensure innovation without undermining creator rights.