TDM, GenAI and the Copyright Three-Step Test

TDM, GenAI and the Copyright Three-Step Test external link

Abstract

In the debate on copyright exceptions permitting text and data mining (“TDM”) for the development of generative AI systems, the so-called “three-step test” has become a centre of gravity. The test serves as a universal yardstick for assessing the compatibility of domestic copyright exceptions with international copyright law. However, it is doubtful whether the international three-step test is applicable at all. Arguably, TDM copies fall outside the scope of the international right of reproduction and go beyond the ambit of the test’s operation. Only if national or regional copyright legislation declares the test applicable, the question arises whether copyright exceptions supporting TDM for AI training constitute certain special cases that do not conflict with a work’s normal exploitation and do not unreasonably prejudice legitimate author or rightsholder interests. As the following analysis will show, rules permitting TDM for AI training can satisfy all test criteria. An opt-out opportunity for copyright owners bans the risk of a conflict with a work’s normal exploitation and an unreasonable prejudice from the outset. A clear focus on specific policy goals, such as the objective to support scientific research, adds conceptual contours that dispel concerns about incompliance. In the case of TDM provisions covering commercial AI development, equitable remuneration regimes can be introduced as a counterbalance to avoid an unreasonable prejudice.

Links

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5373903

Copyright, Generative AI, Text and Data Mining (TDM), three-step test

RIS

TY  - UNPD
AU  - Senftleben, M.
PY  - 
DA  - DATE ERROR: pub_date = 
T1  - TDM, GenAI and the Copyright Three-Step Test
UR  - https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5373903
AB  - In the debate on copyright exceptions permitting text and data mining (“TDM”) for the development of generative AI systems, the so-called “three-step test” has become a centre of gravity. The test serves as a universal yardstick for assessing the compatibility of domestic copyright exceptions with international copyright law. However, it is doubtful whether the international three-step test is applicable at all. Arguably, TDM copies fall outside the scope of the international right of reproduction and go beyond the ambit of the test’s operation. Only if national or regional copyright legislation declares the test applicable, the question arises whether copyright exceptions supporting TDM for AI training constitute certain special cases that do not conflict with a work’s normal exploitation and do not unreasonably prejudice legitimate author or rightsholder interests. As the following analysis will show, rules permitting TDM for AI training can satisfy all test criteria. An opt-out opportunity for copyright owners bans the risk of a conflict with a work’s normal exploitation and an unreasonable prejudice from the outset. A clear focus on specific policy goals, such as the objective to support scientific research, adds conceptual contours that dispel concerns about incompliance. In the case of TDM provisions covering commercial AI development, equitable remuneration regimes can be introduced as a counterbalance to avoid an unreasonable prejudice.
K1  - Copyright
K1  - Generative AI
K1  - Text and Data Mining (TDM)
K1  - three-step test
ER  -

Save .RIS

Bibtex

Working paper{nokey,
	title = {TDM, GenAI and the Copyright Three-Step Test},
	author = {Senftleben, M.},
	year = {},
	date = {DATE ERROR: pub_date = },
	abstract = {In the debate on copyright exceptions permitting text and data mining (“TDM”) for the development of generative AI systems, the so-called “three-step test” has become a centre of gravity. The test serves as a universal yardstick for assessing the compatibility of domestic copyright exceptions with international copyright law. However, it is doubtful whether the international three-step test is applicable at all. Arguably, TDM copies fall outside the scope of the international right of reproduction and go beyond the ambit of the test’s operation. Only if national or regional copyright legislation declares the test applicable, the question arises whether copyright exceptions supporting TDM for AI training constitute certain special cases that do not conflict with a work’s normal exploitation and do not unreasonably prejudice legitimate author or rightsholder interests. As the following analysis will show, rules permitting TDM for AI training can satisfy all test criteria. An opt-out opportunity for copyright owners bans the risk of a conflict with a work’s normal exploitation and an unreasonable prejudice from the outset. A clear focus on specific policy goals, such as the objective to support scientific research, adds conceptual contours that dispel concerns about incompliance. In the case of TDM provisions covering commercial AI development, equitable remuneration regimes can be introduced as a counterbalance to avoid an unreasonable prejudice.},
	keywords = {Copyright, Generative AI, Text and Data Mining (TDM), three-step test},
}

Save .bib

Copyright Liability and Generative AI: What’s the Way Forward? download

Szkalej, K.

Nordic Intellectual Property Law Review, iss. : 1, pp: 92-115, 2025

Abstract

The intersection of copyright liability and generative AI has become one of the most complex and debated issues in the field of copyright law. AI systems have advanced significantly to allow the creation of fantastic new content but they are also capable of producing outputs that evoke, adapt, or recreate content that is protected by copyright law, sparking several infringement proceedings against AI companies, particularly in the US. With this rapid evolution comes the need to re-examine existing legal frameworks and theories. In this contribution, I would like to focus on liability challenges at the output stage of AI content generation and share some insights from Sweden to finally ponder about possible paths forward.

Links

NIR2025nr1_15

Artificial intelligence, Copyright, Generative AI, liability

RIS

TY  - JOUR
AU  - Szkalej, K.
PY  - 2025
DA  - 2025-04-29
T1  - Copyright Liability and Generative AI: What’s the Way Forward?
JO  - Nordic Intellectual Property Law Review
IS  - 1
SP  - 92-115
UR  - https://www.ivir.nl/publications/copyright-liability-and-generative-ai-whats-the-way-forward/nir2025nr1_15/
AB  - The intersection of copyright liability and generative AI has become one of the most complex and debated issues in the field of copyright law. AI systems have advanced significantly to allow the creation of fantastic new content but they are also capable of producing outputs that evoke, adapt, or recreate content that is protected by copyright law, sparking several infringement proceedings against AI companies, particularly in the US. With this rapid evolution comes the need to re-examine existing legal frameworks and theories. In this contribution, I would like to focus on liability challenges at the output stage of AI content generation and share some insights from Sweden to finally ponder about possible paths forward.
K1  - Artificial intelligence
K1  - Copyright
K1  - Generative AI
K1  - liability
ER  -

Save .RIS

Bibtex

Article{nokey,
	title = {Copyright Liability and Generative AI: What’s the Way Forward?},
	author = {Szkalej, K.},
	year = {2025},
	date = {2025-04-29},
	journal = {Nordic Intellectual Property Law Review},
	issue = {1},
	pages = {92-115},
	abstract = {The intersection of copyright liability and generative AI has become one of the most complex and debated issues in the field of copyright law. AI systems have advanced significantly to allow the creation of fantastic new content but they are also capable of producing outputs that evoke, adapt, or recreate content that is protected by copyright law, sparking several infringement proceedings against AI companies, particularly in the US. With this rapid evolution comes the need to re-examine existing legal frameworks and theories. In this contribution, I would like to focus on liability challenges at the output stage of AI content generation and share some insights from Sweden to finally ponder about possible paths forward.},
	keywords = {Artificial intelligence, Copyright, Generative AI, liability},
}

Save .bib

European Copyright Society Opinion on Copyright and Generative AI external link

Dusollier, S., Kretschmer, M., Margoni, T., Mezei, P., Quintais, J. & Rognstad, O.A.

Kluwer Copyright Blog, 2025

Links

https://copyrightblog.kluweriplaw.com/2025/02/07/european-copyright-society-opinion-on-copyright-and-generative-ai/

Copyright, Generative AI

RIS

Save .RIS

Bibtex

Save .bib

Generative AI, Copyright and the AI Act external link

Quintais, J.

Computer Law & Security Review, vol. 56, num: 106107, 2025

Abstract

This paper provides a critical analysis of the Artificial Intelligence (AI) Act's implications for the European Union (EU) copyright acquis, aiming to clarify the complex relationship between AI regulation and copyright law while identifying areas of legal ambiguity and gaps that may influence future policymaking. The discussion begins with an overview of fundamental copyright concerns related to generative AI, focusing on issues that arise during the input, model, and output stages, and how these concerns intersect with the text and data mining (TDM) exceptions under the Copyright in the Digital Single Market Directive (CDSMD). The paper then explores the AI Act's structure and key definitions relevant to copyright law. The core analysis addresses the AI Act's impact on copyright, including the role of TDM in AI model training, the copyright obligations imposed by the Act, requirements for respecting copyright law—particularly TDM opt-outs—and the extraterritorial implications of these provisions. It also examines transparency obligations, compliance mechanisms, and the enforcement framework. The paper further critiques the current regime's inadequacies, particularly concerning the fair remuneration of creators, and evaluates potential improvements such as collective licensing and bargaining. It also assesses legislative reform proposals, such as statutory licensing and AI output levies, and concludes with reflections on future directions for integrating AI governance with copyright protection.

Links

AI Act, Content moderation, Copyright, Digital Services Act (DSA), Generative AI, Text and Data Mining (TDM), Transparency

RIS

TY  - JOUR
AU  - Quintais, J.
PY  - 2025
DA  - 2025-01-30
T1  - Generative AI, Copyright and the AI Act
JO  - Computer Law & Security Review
VL  - 56
NV  - 106107
UR  - https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4912701
DO  - https://doi.org/10.1016/j.clsr.2025.106107
AB  - This paper provides a critical analysis of the Artificial Intelligence (AI) Act\'s implications for the European Union (EU) copyright acquis, aiming to clarify the complex relationship between AI regulation and copyright law while identifying areas of legal ambiguity and gaps that may influence future policymaking. The discussion begins with an overview of fundamental copyright concerns related to generative AI, focusing on issues that arise during the input, model, and output stages, and how these concerns intersect with the text and data mining (TDM) exceptions under the Copyright in the Digital Single Market Directive (CDSMD).
The paper then explores the AI Act\'s structure and key definitions relevant to copyright law. The core analysis addresses the AI Act\'s impact on copyright, including the role of TDM in AI model training, the copyright obligations imposed by the Act, requirements for respecting copyright law—particularly TDM opt-outs—and the extraterritorial implications of these provisions. It also examines transparency obligations, compliance mechanisms, and the enforcement framework. The paper further critiques the current regime\'s inadequacies, particularly concerning the fair remuneration of creators, and evaluates potential improvements such as collective licensing and bargaining. It also assesses legislative reform proposals, such as statutory licensing and AI output levies, and concludes with reflections on future directions for integrating AI governance with copyright protection.
K1  - AI Act
K1  - Content moderation
K1  - Copyright
K1  - Digital Services Act (DSA)
K1  - Generative AI
K1  - Text and Data Mining (TDM)
K1  - Transparency
ER  -

Save .RIS

Bibtex

Article{nokey,
	title = {Generative AI, Copyright and the AI Act},
	author = {Quintais, J.},
	url = {https://www.sciencedirect.com/science/article/pii/S0267364925000020},
	doi = {https://doi.org/10.1016/j.clsr.2025.106107},
	year = {2025},
	date = {2025-01-30},
	journal = {Computer Law & Security Review},
	volume = {56},
	number = {106107},
	abstract = {This paper provides a critical analysis of the Artificial Intelligence (AI) Act\'s implications for the European Union (EU) copyright acquis, aiming to clarify the complex relationship between AI regulation and copyright law while identifying areas of legal ambiguity and gaps that may influence future policymaking. The discussion begins with an overview of fundamental copyright concerns related to generative AI, focusing on issues that arise during the input, model, and output stages, and how these concerns intersect with the text and data mining (TDM) exceptions under the Copyright in the Digital Single Market Directive (CDSMD).
The paper then explores the AI Act\'s structure and key definitions relevant to copyright law. The core analysis addresses the AI Act\'s impact on copyright, including the role of TDM in AI model training, the copyright obligations imposed by the Act, requirements for respecting copyright law—particularly TDM opt-outs—and the extraterritorial implications of these provisions. It also examines transparency obligations, compliance mechanisms, and the enforcement framework. The paper further critiques the current regime\'s inadequacies, particularly concerning the fair remuneration of creators, and evaluates potential improvements such as collective licensing and bargaining. It also assesses legislative reform proposals, such as statutory licensing and AI output levies, and concludes with reflections on future directions for integrating AI governance with copyright protection.},
	keywords = {AI Act, Content moderation, Copyright, Digital Services Act (DSA), Generative AI, Text and Data Mining (TDM), Transparency},
}

Save .bib

Anticipating impacts: using large‑scale scenario‑writing to explore diverse implications of generative AI in the news environment

Kieslich, K., Diakopoulos, N. & Helberger, N.

AI and Ethics, 2024

Abstract

The tremendous rise of generative AI has reached every part of society—including the news environment. There are many concerns about the individual and societal impact of the increasing use of generative AI, including issues such as disinformation and misinformation, discrimination, and the promotion of social tensions. However, research on anticipating the impact of generative AI is still in its infancy and mostly limited to the views of technology developers and/or researchers. In this paper, we aim to broaden the perspective and capture the expectations of three stakeholder groups (news consumers; technology developers; content creators) about the potential negative impacts of generative AI, as well as mitigation strategies to address these. Methodologically, we apply scenario-writing and use participatory foresight in the context of a survey (n=119) to delve into cognitively diverse imaginations of the future. We qualitatively analyze the scenarios using thematic analysis to systematically map potential impacts of generative AI on the news environment, potential mitigation strategies, and the role of stakeholders in causing and mitigating these impacts. In addition, we measure respondents' opinions on a specifc mitigation strategy, namely transparency obligations as suggested in Article 52 of the draft EU AI Act. We compare the results across diferent stakeholder groups and elaborate on diferent expected impacts across these groups. We conclude by discussing the usefulness of scenario-writing and participatory foresight as a toolbox for generative AI impact assessment.

Links

DOI: https://doi.org/10.1007/s43681-024-00497-4

Generative AI

RIS

TY  - GEN
AU  - Kieslich, K.
AU  - Diakopoulos, N.
AU  - Helberger, N.
PY  - 2024
DA  - 2024-05-27
T1  - Anticipating impacts: using large‑scale scenario‑writing to explore diverse implications of generative AI in the news environment
JO  - AI and Ethics
DO  - https://doi.org/10.1007/s43681-024-00497-4
AB  - The tremendous rise of generative AI has reached every part of society—including the news environment. There are many
concerns about the individual and societal impact of the increasing use of generative AI, including issues such as disinformation and misinformation, discrimination, and the promotion of social tensions. However, research on anticipating the
impact of generative AI is still in its infancy and mostly limited to the views of technology developers and/or researchers.
In this paper, we aim to broaden the perspective and capture the expectations of three stakeholder groups (news consumers;
technology developers; content creators) about the potential negative impacts of generative AI, as well as mitigation strategies to address these. Methodologically, we apply scenario-writing and use participatory foresight in the context of a survey
(n=119) to delve into cognitively diverse imaginations of the future. We qualitatively analyze the scenarios using thematic
analysis to systematically map potential impacts of generative AI on the news environment, potential mitigation strategies,
and the role of stakeholders in causing and mitigating these impacts. In addition, we measure respondents\' opinions on a
specifc mitigation strategy, namely transparency obligations as suggested in Article 52 of the draft EU AI Act. We compare
the results across diferent stakeholder groups and elaborate on diferent expected impacts across these groups. We conclude
by discussing the usefulness of scenario-writing and participatory foresight as a toolbox for generative AI impact assessment.
K1  - Generative AI
ER  -

Save .RIS

Bibtex

article{nokey,
	title = {Anticipating impacts: using large‑scale scenario‑writing to explore diverse implications of generative AI in the news environment},
	author = {Kieslich, K. and Diakopoulos, N. and Helberger, N.},
	doi = {https://doi.org/10.1007/s43681-024-00497-4},
	year = {2024},
	date = {2024-05-27},
	journal = {AI and Ethics},
	abstract = {The tremendous rise of generative AI has reached every part of society—including the news environment. There are many
concerns about the individual and societal impact of the increasing use of generative AI, including issues such as disinformation and misinformation, discrimination, and the promotion of social tensions. However, research on anticipating the
impact of generative AI is still in its infancy and mostly limited to the views of technology developers and/or researchers.
In this paper, we aim to broaden the perspective and capture the expectations of three stakeholder groups (news consumers;
technology developers; content creators) about the potential negative impacts of generative AI, as well as mitigation strategies to address these. Methodologically, we apply scenario-writing and use participatory foresight in the context of a survey
(n=119) to delve into cognitively diverse imaginations of the future. We qualitatively analyze the scenarios using thematic
analysis to systematically map potential impacts of generative AI on the news environment, potential mitigation strategies,
and the role of stakeholders in causing and mitigating these impacts. In addition, we measure respondents\' opinions on a
specifc mitigation strategy, namely transparency obligations as suggested in Article 52 of the draft EU AI Act. We compare
the results across diferent stakeholder groups and elaborate on diferent expected impacts across these groups. We conclude
by discussing the usefulness of scenario-writing and participatory foresight as a toolbox for generative AI impact assessment.},
	keywords = {Generative AI},
}

Save .bib

Generative AI and copyright: Convergence of opt-outs? external link

Keller, P.

Links

https://legalblogs.wolterskluwer.com/copyright-blog/generative-ai-and-copyright-convergence-of-opt-outs/

convergence, Copyright, Generative AI, Text and Data Mining (TDM)

RIS

Save .RIS

Bibtex

Save .bib