Student or educator? Sign up with your school email and get 6 months of Pro free.

The DOI matches its Crossref record; 7 of the 7 references it cites were found in a registry or an index; none of the 31 sources its claims cite is checked for support yet.

Rethinking copyright exceptions in the era of generative AI: Balancing innovation and intellectual property protection

Open access
Thongmeensuk, Saliltorn·2024·The Journal of World Intellectual Property

Abstract

Generative artificial intelligence (AI) systems, together with text and data mining (TDM), introduce complex challenges at the junction of data utilization and copyright laws. The inherent reliance of AI on large quantities of data, often encompassing copyrighted materials, results in multifaceted legal quandaries. Issues surface from the unfeasible task of securing permission from each copyright holder for AI training, further muddled by ambiguities in interpreting copyright laws and fair use provisions. Adding to the conundrum, the clandestine practices of data collection in proprietary AI systems obstruct copyright owners from detecting unauthorized use of their materials. The paper explores the exceptions to copyright laws for TDM in the European Union, the United Kingdom, and Japan, recognizing their crucial role in fostering AI development. The EU has a two‐pronged approach under the Directive on Copyright in the Digital Single Market, with one exception catering specifically to research organizations, and another, more generalized one, that can be restricted by rightsholders. The UK allows noncommercial TDM research without infringement but rejected a broader copyright exception due to concerns from the creative sector. Japan has the broadest TDM exception globally, permitting the nonenjoyment use of works without permission, though this can potentially overlook the rights of copyright owners. Notably, the applicability of TDM exceptions to AI‐produced copies remains unclear, creating potential legal challenges. Furthermore, an exploration of the fair use doctrine in the United States provides insight into its potential application in AI development. It focuses on the transformative aspect of usage and its impact on the original work's potential market. This exploration underscores the necessity for clear, practical guidelines. In response to these identified challenges, this paper proposes a hybrid model for TDM exceptions emerges, along with recommended specific mechanisms. The model divides exceptions into noncommercial and commercial uses, providing a nuanced solution to complex copyright issues in AI training. Recommendations incorporate mandatory exceptions for noncommercial uses, an opt‐out clause for commercial uses, enhanced transparency measures, and a searchable portal for copyright owners. In conclusion, striking a delicate equilibrium between technological progress and the incentive for creative expression is of paramount importance. These suggested solutions aim to establish a harmonious foundation that nurtures innovation and creativity while honoring creators' rights, facilitating AI development, promoting transparency, and ensuring fair compensation for creators.

Overview

CoChat Summary

TL;DR

This paper comparatively analyzes text and data mining (TDM) copyright exceptions in the EU, UK, and Japan, plus US fair use and the EU AI Act, to assess how copyright law should balance generative AI innovation with creator protection. It proposes a hybrid TDM model that separates noncommercial and commercial uses, combining a mandatory noncommercial exception with disclosure and opt-out mechanisms for commercial AI training.

Key findings

  • The EU CDSM Directive has two TDM provisions: Article 3 allows TDM for scientific research by research organizations and cultural heritage institutions and is mandatory, while Article 4 is broader but can be reserved by rightsholders through opt-out or contract-out.
  • The UK’s section 29A CDPA 1988 exception permits copying for computational analysis only for noncommercial research and cannot be restricted by contract; after a 2022 consultation, the UK did not proceed with a broader exception because the creative sector raised concerns about licensing and remuneration.
  • Japan’s Article 30-4 is described as the broadest TDM exception globally: it permits commercial and noncommercial TDM for “nonenjoyment” purposes without rightsholder permission or reservation, and without requiring lawful access, though the paper says this may neglect copyright owners’ legitimate interests and challenge the TRIPS “Three-Step” Test.
  • In the US, fair use is a flexible, case-by-case four-factor analysis; AI training copies may be transformative, but outputs that mimic or compete with originals—such as the AI-generated song “Heart on My Sleeve”—can raise fourth-factor market-harm concerns. The EU AI Act adds a requirement for generative AI providers to disclose a “sufficiently detailed” summary of copyright-protected training data.

Methods

  • Comparative legal analysis of copyright frameworks and TDM exceptions in the EU, UK, and Japan, with additional analysis of US fair use doctrine and the EU AI Act.
  • The paper examines statutes, directives, consultations, case law, and secondary commentary, including EU CDSM Articles 3 and 4, UK CDPA section 29A, Japanese Copyright Act Article 30-4, and US fair use cases.
  • It proposes a hybrid model for TDM exceptions and discusses implementation mechanisms such as disclosure, opt-out, searchable portals, and international harmonization through TRIPS, CPTPP, OECD, and UNESCO forums. No new data were created or analyzed.

Limitations

  • Stated: The paper says no new data were created or analyzed, so the analysis is doctrinal and comparative rather than empirical.
  • Stated: It notes unresolved ambiguities, including the EU CDSM Directive’s lack of a clear definition of “scientific research” and the EU AI Act’s undefined “sufficiently detailed” disclosure requirement.
  • Evident from the text: The proposed hybrid model is not empirically tested, and the analysis is limited to selected jurisdictions—EU, UK, Japan, and US fair use—so broader global implementation challenges remain largely theoretical.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1111/jwip.12301

Cites

0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

0 Valid · 5 Not valid · 30 Not checked · 39 Own results and interpretations

19 more claims

Create a free account to see all 74 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 38
  • It may state two separate claims as one. 22

Create a free account to see these claims.

Citations

Cites

The publisher hasn’t deposited this paper’s reference list.

Research works printed but not linked: 8

  1. Andersen v. Stability AI Ltd. (3:23-cv-00201). 17 Ibid. THONGMEENSUK | 291 17471796, 2024, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jwip.12301 by Mount Vernon Nazarene University, Wiley Online Library on [31/12/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 18 Getty Images (US), Inc. v. Stability AI, Inc. (1:23-cv-00135). 19 Ibid. · doi:10.1111/jwip.12301bymountvernonnazareneuniversity
  2. Marti A. Hearst, 'Untangling Text and Data Mining', (Proceedings of the 37th Annual meeting of the Association for Computational Linguistics), quoted in Péter Kollár, 'Mind If I Mine? A Study on the Justification and Sufficiency of Text and Data Mining Exceptions in the European Union' (2021) 3.
  3. Yihao Li and Theresa Beaubouef, 'Data Mining: Concepts, Background and Methods of Integrating Uncertainty in Data Mining' (2010) <https://www.ccsc.org/southcentral/E-Journal/2010/Papers/Yihao%20final%20paper%20CCSC %20for%20submission.pdf> accessed 2 August 2023. 22 Directive 2019/790 of the European Parliament and of the Council of 17 April 2019 on Copyright and Related Rights in the Digital Single Market and Amending Directives 96/9/EC and 2001/29/EC, 2019 O.J. (L 130) 92. 23 Directive (EU) 2019/790, art. 2(1). 24 Directive (EU) 2019/790, art. 2(3) and recital 13.
  4. Kollár, 'Mind if I Mine? A Study on the Justification and Sufficiency of Text and Data Mining Exceptions in the European Union' 18; Christophe Geiger and others, 'Text and Data Mining: Articles 3 and 4 of the Directive 2019/790/EU', 9, in Concepción Saiz García and Raquel Evangelio Llorca (eds), Propiedad Intelectual Y Mercado úNico Digital Europeo, no. 2019-08 (2019): 27-71. 32 Directive (EU) 2019/790, recital 18.
  5. Eleonora Rosati, Copyright in the Digital Single Market: Article-by-Article Commentary to the Provisions of Directive 2019/ 790 (Oxford University Press 2021), 89.
  6. Martin Senftleben and others, 'Ensuring the Visibility and Accessibility of European Creative Content on the World Market: The Need for Copyright Data Improvement in the Light of New Technologies' (2021) SSRN EJ 6.
  7. Shyamkrishna Balganesh, Ng-Loy Wee Loon and Haochen Sun, (eds.), The Cambridge Handbook of Copyright Limitations and Exceptions (Cambridge University Press 2021). Implication' (2020) 70(2) GRURI 148.
  8. Daniel J. Gervais, 'AI Derivatives: The Application to the Derivative Work Right to Literary and Artistic Productions of AI Machines' (2022) 53 SHLR. 71 Ibid. THONGMEENSUK | 293 17471796, 2024, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jwip.12301 by Mount Vernon Nazarene University, Wiley Online Library on [31/12/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License · doi:10.1111/jwip.12301bymountvernonnazareneuniversity

Cited by

See all 36

Metadata

Authors
Thongmeensuk, S.
Affiliations
Tashkent State University of Law (UZ)
Corresponding author
Thongmeensuk, Saliltorn
Venue
The Journal of World Intellectual Property
Publisher
Wiley
Published
2024-04-19
Volume
27
Issue
2
Pages
278-295
ISSN
  • 1422-2213 (ISSN-L)
  • 1747-1796
OpenAlex
W4394976704
Cited by
36
References
0
Citations by year
2026: 19 · 2025: 12 · 2024: 5
FWCI
3.80
Type
journal article
Language
en
Open access
Locations
The Journal of World Intellectual Property (publishedVersion)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Law, AI, and Intellectual Property
  • Social Sciences › Law › Intellectual Property Law
  • Social Sciences › Law › Intellectual Property Rights and Media
Keywords
copyright exceptions, text and data mining, generative AI, AI training data, fair use, transformative use, non-commercial use, opt-out clause, copyright holders' rights

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-11-12, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-12 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Rethinking copyright exceptions in the era of generative AI: Balancing innovation and intellectual property protection. https://app.cochat.ai/papers/10.1111/jwip.12301
The paper
Thongmeensuk, S. (2024). Rethinking copyright exceptions in the era of generative AI: Balancing innovation and intellectual property protection. The Journal of World Intellectual Property. https://doi.org/10.1111/jwip.12301

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.