WhatsApp Channel Join Now
Telegram Group Join Now

Relevance: GS-III (Science & Technology, AI, IPR); GS-II (Governance) Source: Tech & Legal Policy Updates, 2026

1 · What is the issue?

Artificial Intelligence (AI) needs massive amounts of high-quality written data to become “smart.” Because the internet is now filled with low-quality, repetitive content, AI companies desperately need traditional, well-edited physical books to train their models.
To bypass digital copyright locks on e-books, companies are doing something shocking: they are buying thousands of physical books in bulk, chopping off their spines, and running the loose pages through high-speed scanners to digitize the text. This process permanently destroys the book. This desperate grab for data has sparked a massive legal debate over whether destroying and scanning copyrighted books without the author’s permission is legal under Indian law.

2 · Understanding the Book Scanning Process

Step 1: The Data Need
AI models learn by analyzing long, complex, and grammatically correct sentences. Physical books offer this clean, vetted data far better than random internet websites.
▼
Step 2: Non-Destructive Scanning (The Old Way)
Normally, books are scanned gently from above while keeping the spine intact. This protects the book but is very slow and often creates blurry pages.
▼
Step 3: Destructive Scanning (The AI Way)
AI companies cut off the book’s spine so the loose pages can be fed into a rapid, automated scanner. The book is destroyed, but the scan is perfect and lightning-fast.
▼
Step 4: The Legal Argument
Authors argue this is massive copyright theft. Tech companies argue that because the AI only learns patterns (and doesn’t just copy-paste the book), it should be perfectly legal.

3 · Key Legal Provisions

Section 52
Fair Dealing
Under India’s Copyright Act (1957), you are allowed to use small parts of a copyrighted work without permission if it is for “private study, research, or review.” AI companies claim training AI counts as research.
Non-Expressive Use
Not Copying
This legal argument states that AI isn’t reading the book to copy the story. It is only performing math (tokenization) to learn grammar patterns, which shouldn’t break copyright laws.
Delhi HC Ruling (2026)
A Win for AI
In a massive recent case, the Delhi High Court ruled that OpenAI training ChatGPT on copyrighted news articles is currently legal under the “Fair Dealing” research rule.
Proposed Solution
Blanket Licenses
The government has proposed a “One Nation, One License” rule. AI companies can legally scan whatever they want for training, but they must automatically pay royalties to the original authors.

UPSC Prelims Quick Facts: Tech & Law
AI “Slop” AI slop is a term for low-quality, mass-produced digital content created using generative artificial intelligence. It floods social media feeds, search engines, and streaming platforms with minimal human effort or genuine meaning, often designed purely to chase clicks, views, or ad revenue.
IndiaAI Mission India’s official national mission aiming to build purely indigenous, locally trained “Sovereign AI” models to reduce dependence on foreign tech.
Copyright Act 1957 The old Indian law that is currently struggling to regulate modern AI because it was written decades before digital machine learning existed.
Dataset Transparency Experts are pushing for mandatory “transparency obligations,” forcing AI companies to clearly reveal exactly which books they used to train their models.

MCQ Practice Question
Q. With reference to Intellectual Property Rights (IPR) and Artificial Intelligence in India, consider the following statements:

  1. Under Section 52 of the Copyright Act, 1957, limited use of copyrighted material without permission is permitted strictly for personal research and study under the “Fair Dealing” exception.
  2. In a landmark 2026 ruling, the Delhi High Court declared that using copyrighted news content to train Large Language Models like ChatGPT is completely illegal and constitutes criminal copyright infringement.
  3. “Destructive scanning” is an automated digitization process favoured by AI companies because it protects the physical integrity of fragile books while copying their text.

Which of the statements given above is/are correct?
(a) 1 only    (b) 1 and 2 only    (c) 2 and 3 only    (d) 1, 2 and 3

Answer: (a) 1 only

  • Statement 1 — Correct: Section 52 establishes the “Fair Dealing” exception, legally permitting the use of copyrighted material for purposes like private study, research, and review.
  • Statement 2 — Incorrect: The Delhi High Court actually ruled in favor of AI, stating that scraping news content to train LLMs falls under the “fair dealing” (research) exception and refused to ban OpenAI.
  • Statement 3 — Incorrect (the trap): Destructive scanning permanently destroys the physical book by cutting off the spine to feed the loose pages into rapid automated scanners.
Ajmal IAS APSC Foundation Course

Recent Posts

Ajmal IAS APSC Foundation BATCH
Ajmal IAS APSC Foundation Course

Start Yours at Ajmal IAS – with Mentorship StrategyDisciplineClarityResults that Drives Success

Your dream deserves this moment — begin it here.