AI, Copyrights, and Libraries

Summary

This article explores the evolving relationship between libraries, copyright law, and artificial intelligence in India. It explains the primary role of libraries, the purpose of copyright law, and the natural conflict between access and control. It outlines copyright rights, infringement, and library-specific exceptions under Indian law, examines emerging AI copyright issues such as training data use, authorship of AI outputs, and infringement by AI-generated works, and highlights how AI can transform libraries through smarter cataloguing, discovery, and access. Finally, it offers practical steps for libraries to adopt AI responsibly through policy frameworks, risk assessments, vendor contracts, and monitoring.

Libraries have always been at the heart of knowledge dissemination. They are repositories of history, ideas, and culture, and they provide access to education, information, and knowledge to the public. In India, libraries have been playing a vital role in advancing literacy and education .
As technology transforms learning, artificial intelligence is bringing powerful tools that can change how libraries collect, manage, and deliver information. However, copyright law still governs how creative works may be used. This creates a natural tension between libraries, which exist to enable access, and copyright law, which regulates that access through a system of exclusive rights. Understanding this tension, and its legal and practical implications, is essential for libraries integrating AI into their operations.

I. Purpose and Role of Libraries

Access to information, knowledge, education, and culture is at the core of what libraries aim to promote. Their role is to collect, organise, preserve, and make available books, journals, manuscripts, films, sound recordings, and now, digital content.

Public libraries make knowledge accessible to the general public, and academic libraries facilitate education, provide access to information and research material, and support scholarship. Digital libraries and online repositories such as the National Digital Library of India, JSTOR, and Project Gutenberg have been extending library services to remote users and enabled round-the-clock access.

II. Copyright Law: Purpose and Objectives

The objective of copyright law is to promote the creation of works by granting authors exclusive rights over their creations. These rights allow copyright owners to control reproduction, distribution, communication, performance, and adaptation of their works. The belief is that these exclusive rights incentivise creators, and enrich the pool of knowledge and culture for society’s benefit.

This grant of rights, however, also restricts access. Users cannot freely copy, share, or distribute works. Libraries, which exist to enable public access, must therefore operate within the framework of copyrights, either by obtaining licences or relying on legal exceptions.

III. The Conflict between Libraries and Copyrights

Promoting maximum access is a foundational aim for libraries. Copyright law, by design, restricts use and distribution to ensure that authors benefit from their works. This creates an unavoidable conflict.

For example, when libraries wish to digitise their collections or create online course packs, they must ensure that these activities are either permitted by law or licensed. In India, where students and researchers often cannot afford expensive textbooks or journals, this conflict becomes sharper.

The Delhi High Court’s decision in the Rameshwari photocopy case illustrates this tension. Publishers sued a photocopy shop for preparing course packs using books from the Delhi University library. The court held that preparing and selling course packs was permissible for educational purposes, and that libraries may make education-related content available through course packs and copying facilities so long as it is within the framework of education and instruction.

IV. Copyrightable Works, Rights, and Infringement

The rights granted include the right to reproduce the work, issue copies to the public, adapt it (including translation), distribute, communicate to the public, and perform it in public.

Any unauthorised exercise of these rights amounts to infringement unless covered by an exception.

Some instances that may give rise to infringement include:

Scanning and circulating an entire book over WhatsApp can amount to infringement.
Playing a film in a library auditorium without a public performance licence may amount to infringement.
Uploading journal articles to an open website without authorisation can give rise to infringement.

To avoid infringement, libraries must assess copyright risks in their actions, and mitigate them.

V. Library Exceptions under Indian Law

Indian copyright law provides specific library exceptions under Section 52 of the Copyright Act, 1957. These exceptions are critical for libraries to function lawfully. Important exceptions include:

Educational Use

Libraries and teachers may use works for teaching, scholarship, and research. This includes reproduction before, during, and after class, as well as preparation of course packs. The Rameshwari case has confirmed that course packs for educational or instructional use are lawful.

Storage / Copy

It is permissible to store electronic copies of works or make copies for preservation and for collection or archival purposes. This includes works for which physical copies are held, unpublished works, and those not available in India.

Fair Dealing / Accessibility

Reading access, extraction, and reproduction for personal research or private use is allowed. Works may also be converted into accessible formats for persons with disabilities and made available accordingly.
These exceptions enable lawful use by libraries, but do not permit large-scale unauthorised commercial exploitation.

VI. AI and Copyright – Current Debates

In today’s context, artificial intelligence is creating new copyright challenges. Three issues dominate the debate:

The first issue is regarding the use of copyrighted works for training AI models. AI systems are trained on massive datasets that often include copyrighted books, articles, images, and music. Whether such training is permissible or infringing is still unresolved.

Second issue relates to ownership of AI-generated works. Courts and copyright offices around the world are considering whether works created entirely by AI are eligible for copyright. The Stephen Thaler cases have thus far rejected claims over machine-generated works.

Third issue concerns infringement by AI outputs. AI may generate text, music, or images that are substantially similar to copyrighted works or mimic the style of authors and artists. This may amount to infringement. Ongoing cases against Bard, Anthropic, Cadre, and Meta, along with Getty Images v. Stability AI and New York Times v. OpenAI, are testing the application of copyright law to AI training and outputs.

In this backdrop, libraries must be mindful and aware of the fact that using AI tools could expose them to liability if outputs are infringing or if training data uses copyrighted material without permission.

VII. Value of AI for Libraries

AI offers immense opportunities for improving services and outreach, including:

Smart cataloguing and metadata tagging
AI-driven recommendation systems
Chatbots and virtual reference services
Advanced search and discovery tools
Digitisation and preservation of rare works
Personalised content delivery
Accessible format conversion for persons with disabilities
Facilitation of course packs, research, and education Libraries worldwide are already adopting AI in innovative ways, and Indian libraries are beginning to follow suit.

VIII. Copyright Liability for Libraries

Using AI The copyright liability of public and academic libraries in the context of AI may not be the same as that of commercial entities. Courts may take into account their nonprofit and educational role and treat them more leniently than profit-making platforms.

That said, risks remain:

Training AI models on copyrighted works without authorisation can attract infringement claims
AI outputs that reproduce or closely imitate protected works may drag libraries into litigation
AI tools that allow large-scale copying or downloading may exceed the scope of copyright exceptions Given that libraries hold much of the world’s copyrighted content, it is only a matter of time before they are involved in copyright disputes related to AI.

IX. Steps Libraries Can Take

AI related copyright risks can be mitigated by libraries by adopting practical safeguards, including:

Creating well-defined policies and guidelines for AI adoption and use
Conducting training and awareness programmes for staff and users on AI and copyright compliance
Using AI only within the framework of copyright exceptions
Performing copyright risk assessments before adopting AI tools
Including clear provisions in contracts and licences with AI vendors and publishers Clearly allocating responsibility for infringement
Regularly auditing and monitoring AI systems to ensure continued compliance

Conclusion

Artificial intelligence gives libraries an unprecedented opportunity to further their function of enabling access to information, knowledge, education, and culture. With the right tools, they can do so more quickly, effectively, and at a much larger scale — and serve students, researchers, and the public at large.

Public and academic libraries occupy a unique position in the copyright and AI debate. They are nonprofit institutions serving a public mission rather than commercial enterprises seeking profit. Many of their activities relating to storage, research, private use, accessibility, education, and instruction are already protected under copyright exceptions.

Nonetheless, libraries are not immune from risk. AI use could still trigger disputes where outputs infringe copyrighted works or where training data includes protected material without consent.

The way forward is careful adoption. A well-defined policy framework, copyright risk assessments, strong contract provisions with vendors, and continuous monitoring will help libraries avoid unnecessary litigation while confidently using AI to make knowledge more open, equitable, and impactful.

References

IFLA, The Role of Libraries in Lifelong Learning, https://www.ifla.org/publications/the-role-of-libraries-in-lifelong-learning/ (last visited Sept 19th. 2025).
WIPO, Understanding Copyright and Related Rights, https://www.wipo.int/edocs/pubdocs/en/wipo_pub_909_2016.pdf (last visited Sept 19th. 2025).
The Copyright Act, No. 14 of 1957 (as last amended in 2012).
The Chancellor, Masters & Scholars of the Univ. of Oxford v. Rameshwari Photocopy Servs., Delhi High Court (Dec. 9, 2016), https://indiankanoon.org/doc/18745008/.
Getty Images (US), Inc. v. Stability AI, Inc., No. 1:23-cv-00135 (D. Del. filed Feb. 2023).
New York Times Co. v. OpenAI, Inc., No. 1:23-cv-11195 (S.D.N.Y. filed Dec. 2023).
Thaler v. Perlmutter, No. 1:22-cv-01564 (D.D.C. 2023).
National Digital Library of India, https://ndl.iitkgp.ac.in/ (last visited Sept 19th. 2025).
Library of Congress AI Projects, https://labs.loc.gov/work/experiments/machine-learning/ (last visited Sept 19th. 2025).