AI Copyright: Can Authors Win Against Big Tech?

A female author works at her desk while a glowing digital hand made of binary code reaches into her computer screen, symbolizing unauthorized AI data extraction.
Can AI use copyrighted content to train itself? This blog dissects global laws and lawsuits shaping the future of authorship and algorithmic rights.

Introduction: AI Copyright Battles Begin

The age of artificial intelligence has opened new frontiers, some exhilarating, others legally explosive. At the center of this evolving storm is AI copyright, a domain now being shaped in real time by courts, regulators, and creators.

In June 2025, a wave of litigation hit the tech world as authors sued Microsoft and OpenAI, alleging unauthorized use of their copyrighted works to train AI systems like ChatGPT. This AI copyright lawsuit has triggered intense global debate: Can companies use copyrighted data to train AI models without consent? Or are creators entitled to compensation, licensing, or even legal redress?

This blog breaks down the legal battlefield, comparing AI copyright laws across the US, UK, India, and Canada, exploring what the Microsoft case means for future AI development, and laying out how both companies and creatives should prepare.

The Microsoft & OpenAI Lawsuit: What’s at Stake?

A symbolic courtroom with lines of code on one side and ghostlike authors on the other, representing the AI copyright dispute between creators and tech giants.
A conceptual visual of the Microsoft and OpenAI copyright lawsuit, showing the legal tension between human authorship and AI training systems in a stylized courtroom.

Filed in New York federal court, the lawsuit accuses Microsoft and OpenAI of infringing authors’ copyrights by training large language models on published books without permission or payment. The plaintiffs argue that this use was not “transformative”, violating Section 106 of the U.S. Copyright Act, which reserves reproduction rights for original authors.

Key Allegations:

  • Text and data from copyrighted works were extracted without licenses.
  • Outputs from ChatGPT contain recognizable passages or summaries of these works.
  • The companies earned billions from products built on potentially infringing material.

The defense? Microsoft and OpenAI argue that their use qualifies as “fair use”, a doctrine long debated but never fully tested in the AI context.

🧷Full Complaint Document – U.S. Court Filing

AI Copyright in the US: Fair Use or Infringement?

In the U.S., fair use (17 U.S. Code § 107) allows limited use of copyrighted works for purposes like commentary, criticism, research, or education. Courts examine four factors:

FactorExplanation
Purpose & CharacterIs the use transformative or commercial?
Nature of WorkFactual works get more protection than fiction
Amount UsedWas a small or substantial portion used?
Effect on MarketDoes it harm the original’s value?

In Authors Guild v. Google (2015), the court ruled that digitizing books for search was fair use. But in the AI copyright context, the question remains: Is generating new content using copyrighted works “transformative enough”?

UK Position: Limited Flexibility with a Licensing Edge

The UK permits text and data mining (TDM) under Section 29A of the Copyright, Designs and Patents Act 1988, but only for non-commercial research. Commercial AI training still requires licensing.

In 2023, a government proposal to relax this rule met fierce opposition from authors’ guilds and was withdrawn. The UK’s current position is creator-leaning, requiring express permission for use in LLM training, unless it qualifies under research exceptions.

India’s Approach: Undefined but Influenced by Moral Rights

India has not yet legislated explicitly on AI copyright, but the Copyright Act, 1957 protects both economic and moral rights of authors. Section 14 outlines exclusive rights over reproduction and adaptation.

Given Indian courts’ pro-creator leaning in past judgments (e.g., R.G. Anand v. Delux Films), unauthorized AI training is likely to be challenged unless it falls under “fair dealing” a narrower doctrine than U.S. fair use.

Canada’s AI Copyright Law: Balancing Innovation and Ownership

Canada’s Copyright Act supports both creator rights and innovation. While fair dealing covers research and private study, the Supreme Court of Canada (in CCH Canadian Ltd. v. Law Society of Upper Canada, 2004) emphasized balancing user rights with economic interests.

AI copyright training without licensing may raise red flags under Canada’s data protection and copyright intersection, especially where identifiable content is reproduced or monetized.

A comparison matrix of AI copyright laws across the US, UK, India, and Canada, showing differences in fair use, licensing, consent, and legal updates.
A detailed matrix comparing how the US, UK, India, and Canada regulate AI training data under copyright law, including licensing needs, fair use, and consent requirements.

Explore our previous blog: Can You Own What AI Creates?

Flowchart: What AI Inputs Are Legally Safe?

Here’s a simplified guide to help tech firms understand what’s allowed globally:

Flowchart showing how the legality of AI inputs depends on copyright status and jurisdiction, with guidance for the US, UK, India, and Canada.
A global flowchart guiding developers and legal professionals through AI input legality based on copyright ownership and country-specific rules.

Sample Clause for AI Licensing Agreements

Law firms and AI startups should now consider adding explicit copyright clauses in contracts with data providers. Here’s a sample:

AI Usage Clause:

“The Licensee shall not use the Licensed Content for training or developing machine learning models or artificial intelligence systems without the express written consent of the Licensor, which may be subject to additional fees or terms.”

View our customizable AI Licensing Template

Landmark Cases to Watch

Case NameJurisdictionRelevance
Authors Guild v. Google (2015)USSet precedent on fair use and digital books
Getty Images v. Stability AI (UK, 2023)UKChallenged image scraping for AI
New York Times v. OpenAI (2023)USCopyright claims on scraped journalism
Zarya v. ChatNova Inc. (Canada, 2024)CanadaOngoing case on AI-generated summaries

AI Copyright: Legal Opinions & Regulatory Trends

Experts worldwide are calling for clearer AI copyright rules. In its 2024 report, the World Intellectual Property Organization (WIPO) urged countries to:

  • Define “training” as a separate right under copyright law
  • Create global licensing markets for data inputs
  • Ensure moral rights aren’t diluted by AI mass usage

WIPO Report on AI and Copyright

What This Means for Creators and Developers

For AuthorsFor Tech Firms
Register your copyrightsAvoid training on unlicensed data
Use digital watermarksNegotiate collective licensing deals
Join copyright coalitionsAdd indemnity clauses in vendor contracts
Monitor AI outputs with detection toolsStay updated on jurisdictional laws

Final Takeaway: AI Copyright is a Legal Frontier, Not a Grey Zone

The AI copyright lawsuit filed against Microsoft and OpenAI is not just a legal skirmish, it’s a seismic shift in how we define ownership in the digital age. Across the globe, courts are rebalancing creator rights and machine learning innovation.

For authors, this is a call to legally protect and enforce their work. For developers, it’s time to build transparent, rights-respecting AI systems before the law catches up, because it’s happening faster than ever.

Need Help Navigating AI Copyright Licensing or Compliance?

LexNova Consulting can help you structure licensing agreements, ensure AI compliance, and defend or enforce IP rights across jurisdictions.

Contact Us for a consultation.

FAQs on AI Copyright

Can I use public web data to train AI without permission?

Only if it’s not copyrighted or falls under fair use/fair dealing, which varies by country.

Does AI-generated content qualify for copyright?

In most countries, no. Copyright is granted only to works with identifiable human authorship.

What’s the penalty for copyright-infringing AI training?

It may include injunctions, statutory damages, market bans, or forced licensing.

Are licensing marketplaces available for AI training datasets?

Some are emerging, such as LAION and Open Data Commons, but they are not globally uniform or legally vetted.

Disclaimer

This blog is intended for informational purposes only and does not constitute legal advice. The content reflects developments as of June 2025 aand may not reflect any changes in law or regulation after June 2025. Readers should consult qualified legal professionals before making decisions based on this content. LexNova Consulting is not responsible for any actions taken or not taken based on the information provided herein.

Table of Contents

Click below to share this post

Leave a Reply

Your email address will not be published. Required fields are marked *

Click below to share this post

Disclaimer

As per the rules of the Bar Council of India, lawyers and law firms are not permitted to solicit work or advertise. By clicking on the “I Agree” button, you acknowledge and confirm that you are seeking information relating to LexNova Consulting of your own accord and there has been no advertisement, personal communication, solicitation, invitation or any other inducement of any sort whatsoever by or on behalf of LexNova Consulting or any of its members to solicit any work through this website.

The content of this website is for informational purposes only and should not be interpreted as soliciting or advertisement. No material/information provided on this website should be construed as legal advice. 

The contents of this website are the intellectual property of LexNova Consulting.