Generative AI and Creative Commons Licences – The Application of Share Alike Obligations to Trained Models, Curated Datasets and AI Output external link

JIPITEC, vol. 15, iss. : 3, 2024

Abstract

This article maps the impact of Share Alike (SA) obligations and copyleft licensing on machine learning, AI training, and AI-generated content. It focuses on the SA component found in some of the Creative Commons (CC) licences, distilling its essential features and layering them onto machine learning and content generation workflows. Based on our analysis, there are three fundamental challenges related to the life cycle of these licences: tracing and establishing copyright-relevant uses during the development phase (training), the interplay of licensing conditions with copyright exceptions and the identification of copyright-protected traces in AI output. Significant problems can arise from several concepts in CC licensing agreements (‘adapted material’ and ‘technical modification’) that could serve as a basis for applying SA conditions to trained models, curated datasets and AI output that can be traced back to CC material used for training purposes. Seeking to transpose Share Alike and copyleft approaches to the world of generative AI, the CC community can only choose between two policy approaches. On the one hand, it can uphold the supremacy of copyright exceptions. In countries and regions that exempt machine-learning processes from the control of copyright holders, this approach leads to far-reaching freedom to use CC resources for AI training purposes. At the same time, it marginalises SA obligations. On the other hand, the CC community can use copyright strategically to extend SA obligations to AI training results and AI output. To achieve this goal, it is necessary to use rights reservation mechanisms, such as the opt-out system available in EU copyright law, and subject the use of CC material in AI training to SA conditions. Following this approach, a tailor-made licence solution can grant AI developers broad freedom to use CC works for training purposes. In exchange for the training permission, however, AI developers would have to accept the obligation to pass on – via a whole chain of contractual obligations – SA conditions to recipients of trained models and end users generating AI output.

ai, Copyright, creative commons, Licensing, machine learning

Bibtex

Article{nokey, title = {Generative AI and Creative Commons Licences – The Application of Share Alike Obligations to Trained Models, Curated Datasets and AI Output}, author = {Szkalej, K. and Senftleben, M.}, url = {https://www.jipitec.eu/jipitec/article/view/415}, year = {2024}, date = {2024-12-13}, journal = {JIPITEC}, volume = {15}, issue = {3}, pages = {}, abstract = {This article maps the impact of Share Alike (SA) obligations and copyleft licensing on machine learning, AI training, and AI-generated content. It focuses on the SA component found in some of the Creative Commons (CC) licences, distilling its essential features and layering them onto machine learning and content generation workflows. Based on our analysis, there are three fundamental challenges related to the life cycle of these licences: tracing and establishing copyright-relevant uses during the development phase (training), the interplay of licensing conditions with copyright exceptions and the identification of copyright-protected traces in AI output. Significant problems can arise from several concepts in CC licensing agreements (‘adapted material’ and ‘technical modification’) that could serve as a basis for applying SA conditions to trained models, curated datasets and AI output that can be traced back to CC material used for training purposes. Seeking to transpose Share Alike and copyleft approaches to the world of generative AI, the CC community can only choose between two policy approaches. On the one hand, it can uphold the supremacy of copyright exceptions. In countries and regions that exempt machine-learning processes from the control of copyright holders, this approach leads to far-reaching freedom to use CC resources for AI training purposes. At the same time, it marginalises SA obligations. On the other hand, the CC community can use copyright strategically to extend SA obligations to AI training results and AI output. To achieve this goal, it is necessary to use rights reservation mechanisms, such as the opt-out system available in EU copyright law, and subject the use of CC material in AI training to SA conditions. Following this approach, a tailor-made licence solution can grant AI developers broad freedom to use CC works for training purposes. In exchange for the training permission, however, AI developers would have to accept the obligation to pass on – via a whole chain of contractual obligations – SA conditions to recipients of trained models and end users generating AI output.}, keywords = {ai, Copyright, creative commons, Licensing, machine learning}, }

The paradox of lawful text and data mining? Some experiences from the research sector and where we (should) go from here external link

Abstract

Scientific research can be tricky business. This paper critically explores the 'lawful access' requirement in European copyright law which applies to text and data mining (TDM) carried out for the purpose of scientific research. Whereas TDM is essential for data analysis, artificial intelligence (AI) and innovation, the paper argues that the 'lawful access' requirement in Article 3 CDSM Directive may actually restrict research by complicating the applicability of the TDM provision or even rendering it inoperable. Although the requirement is intended to ensure that researchers act in good faith before deploying TMD tools for purposes such as machine learning, it forces them to ask for permission to access data, for example by taking out a subscription to a service, and for that reason provides the opportunity for copyright holders to apply all sorts of commercial strategies to set the legal and technological parameters of access and potentially even circumvent the mandatory character of the provision. The paper concludes by drawing on insights from the recent European Commission study 'Improving access to and reuse of research results, publications and data for scientific purposes' that offer essential perspectives for the future of TDM, and by suggesting a number of paths forward that EU Member States can take already now in order to support a more predictable and reliable legal regime for scientific TDM and potentially code mining to foster innovation.

ai, CDSM Directive, Copyright, text and data mining

Bibtex

Article{nokey, title = {The paradox of lawful text and data mining? Some experiences from the research sector and where we (should) go from here}, author = {Szkalej, K.}, url = {https://ssrn.com/abstract=5000116 }, doi = {https://doi.org/10.2139/ssrn.5000116 }, year = {2024}, date = {2024-11-04}, abstract = {Scientific research can be tricky business. This paper critically explores the \'lawful access\' requirement in European copyright law which applies to text and data mining (TDM) carried out for the purpose of scientific research. Whereas TDM is essential for data analysis, artificial intelligence (AI) and innovation, the paper argues that the \'lawful access\' requirement in Article 3 CDSM Directive may actually restrict research by complicating the applicability of the TDM provision or even rendering it inoperable. Although the requirement is intended to ensure that researchers act in good faith before deploying TMD tools for purposes such as machine learning, it forces them to ask for permission to access data, for example by taking out a subscription to a service, and for that reason provides the opportunity for copyright holders to apply all sorts of commercial strategies to set the legal and technological parameters of access and potentially even circumvent the mandatory character of the provision. The paper concludes by drawing on insights from the recent European Commission study \'Improving access to and reuse of research results, publications and data for scientific purposes\' that offer essential perspectives for the future of TDM, and by suggesting a number of paths forward that EU Member States can take already now in order to support a more predictable and reliable legal regime for scientific TDM and potentially code mining to foster innovation.}, keywords = {ai, CDSM Directive, Copyright, text and data mining}, }

Prompts tussen vorm en inhoud: de eerste rechtspraak over generatieve AI en het werk download

Auteursrecht, iss. : 3, pp: 129-134, 2024

Abstract

Kan het gebruik van generatieve AI-systemen een auteursrechtelijk beschermd werk opleveren? Twee jaar na de introductie van Dall-E en ChatGPT begint zich enige jurisprudentie te vormen. Daarbij is de kernvraag of het aansturen van dergelijke systemen door middel van prompts (instructies) voldoende is om de output als ‘werk’ te kwalificeren. Dit artikel gaat, mede aan de hand van de vroegste rechtspraak in de Verenigde Staten, China en Europa, dieper in op deze lastige kwestie.

ai, Copyright

Bibtex

Article{nokey, title = {Prompts tussen vorm en inhoud: de eerste rechtspraak over generatieve AI en het werk}, author = {Hugenholtz, P.}, url = {https://www.ivir.nl/nl/publications/prompts-tussen-vorm-en-inhoud-de-eerste-rechtspraak-over-generatieve-ai-en-het-werk/auteursrecht2024_3/}, year = {2024}, date = {2024-08-30}, journal = {Auteursrecht}, issue = {3}, abstract = {Kan het gebruik van generatieve AI-systemen een auteursrechtelijk beschermd werk opleveren? Twee jaar na de introductie van Dall-E en ChatGPT begint zich enige jurisprudentie te vormen. Daarbij is de kernvraag of het aansturen van dergelijke systemen door middel van prompts (instructies) voldoende is om de output als ‘werk’ te kwalificeren. Dit artikel gaat, mede aan de hand van de vroegste rechtspraak in de Verenigde Staten, China en Europa, dieper in op deze lastige kwestie.}, keywords = {ai, Copyright}, }

The commodification of trust external link

Blockchain & Society Policy Research Lab Research Nodes, num: 1, 2021

Abstract

Fundamental, wide-ranging, and highly consequential transformations take place in interpersonal, and systemic trust relations due to the rapid adoption of complex, planetary-scale digital technological innovations. Trust is remediated by planetary scale techno-social systems, which leads to the privatization of trust production in society, and the ultimate commodification of trust itself. Modern societies rely on communal, public and private logics of trust production. Communal logics produce trust by the group for the group, and are based on familiar, ethnic, religious or tribal relations, professional associations epistemic or value communities, groups with shared location or shared past. Public trust logics developed in the context of the modern state, and produce trust as a free public service. Abstract, institutionalized frameworks, institutions, such as the press, or public education, science, various arms of the bureaucratic state create familiarity, control, and insurance in social, political, and economic relations. Finally, private trust producers sell confidence as a product: lawyers, accountants, credit rating agencies, insurers, but also commercial brands offer trust for a fee. With the emergence of the internet and digitization, a new class of private trust producers emerged. Online reputation management services, distributed ledgers, and AI-based predictive systems are widely adopted technological infrastructures, which are designed to facilitate trust-necessitating social, economic interactions by controlling the past, the present and the future, respectively. These systems enjoy immense economic success, and they are adopted en masse by individuals and institutional actors alike. The emergence of the private, technical means of trust production paves the way towards the widescale commodification of trust, where trust is produced as a commercial activity, conducted by private parties, for economic gain, often far removed from the loci where trust-necessitating social interactions take place. The remediation and consequent privatization and commodification of trust production has a number of potentially adverse social effects: it may decontextualize trust relationships; it removes trust from the local social, cultural relational contexts; it changes the calculus of interpersonal trust relations. Maybe more importantly as more and more social and economic relations are conditional upon having access to, and good standing in private trust infrastructures, commodification turns trust into the question of continuous labor, or devastating exclusion. By invoking Karl Polanyi’s work on fictious commodities, I argue that the privatization, and commodification of trust may have a catastrophic impact on the most fundamental layers of the social fabric.

ai, blockchains, commodification, frontpage, Informatierecht, Karl Polanyi, reputation, trust, trust production

Bibtex

Article{Bodó2021, title = {The commodification of trust}, author = {Bodó, B.}, url = {https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3843707}, year = {0517}, date = {2021-05-17}, journal = {Blockchain & Society Policy Research Lab Research Nodes}, number = {1}, abstract = {Fundamental, wide-ranging, and highly consequential transformations take place in interpersonal, and systemic trust relations due to the rapid adoption of complex, planetary-scale digital technological innovations. Trust is remediated by planetary scale techno-social systems, which leads to the privatization of trust production in society, and the ultimate commodification of trust itself. Modern societies rely on communal, public and private logics of trust production. Communal logics produce trust by the group for the group, and are based on familiar, ethnic, religious or tribal relations, professional associations epistemic or value communities, groups with shared location or shared past. Public trust logics developed in the context of the modern state, and produce trust as a free public service. Abstract, institutionalized frameworks, institutions, such as the press, or public education, science, various arms of the bureaucratic state create familiarity, control, and insurance in social, political, and economic relations. Finally, private trust producers sell confidence as a product: lawyers, accountants, credit rating agencies, insurers, but also commercial brands offer trust for a fee. With the emergence of the internet and digitization, a new class of private trust producers emerged. Online reputation management services, distributed ledgers, and AI-based predictive systems are widely adopted technological infrastructures, which are designed to facilitate trust-necessitating social, economic interactions by controlling the past, the present and the future, respectively. These systems enjoy immense economic success, and they are adopted en masse by individuals and institutional actors alike. The emergence of the private, technical means of trust production paves the way towards the widescale commodification of trust, where trust is produced as a commercial activity, conducted by private parties, for economic gain, often far removed from the loci where trust-necessitating social interactions take place. The remediation and consequent privatization and commodification of trust production has a number of potentially adverse social effects: it may decontextualize trust relationships; it removes trust from the local social, cultural relational contexts; it changes the calculus of interpersonal trust relations. Maybe more importantly as more and more social and economic relations are conditional upon having access to, and good standing in private trust infrastructures, commodification turns trust into the question of continuous labor, or devastating exclusion. By invoking Karl Polanyi’s work on fictious commodities, I argue that the privatization, and commodification of trust may have a catastrophic impact on the most fundamental layers of the social fabric.}, keywords = {ai, blockchains, commodification, frontpage, Informatierecht, Karl Polanyi, reputation, trust, trust production}, }

News Recommenders and Cooperative Explainability: Confronting the contextual complexity in AI explanations external link

ai, frontpage, news recommenders, Technologie en recht

Bibtex

Report{Drunen2020b, title = {News Recommenders and Cooperative Explainability: Confronting the contextual complexity in AI explanations}, author = {Drunen, M. van and Ausloos, J. and Appelman, N. and Helberger, N.}, url = {https://www.ivir.nl/publicaties/download/Visiepaper-explainable-AI-final.pdf}, year = {1103}, date = {2020-11-03}, keywords = {ai, frontpage, news recommenders, Technologie en recht}, }

Netherlands/Research external link

1029, pp: 164-175

Abstract

How are AI-based systems being used by private companies and public authorities in Europe? The new report by AlgorithmWatch and Bertelsmann Stiftung sheds light on what role automated decision-making (ADM) systems play in our lives. As a result of the most comprehensive research on the issue conducted in Europe so far, the report covers the current use of and policy debates around ADM systems in 16 European countries and at EU level.

ai, automated decision making, frontpage, Technologie en recht

Bibtex

Chapter{Fahy2020b, title = {Netherlands/Research}, author = {Fahy, R. and Appelman, N.}, url = {https://www.ivir.nl/publicaties/download/Automating-Society-Report-2020.pdf https://automatingsociety.algorithmwatch.org/}, year = {1029}, date = {2020-10-29}, abstract = {How are AI-based systems being used by private companies and public authorities in Europe? The new report by AlgorithmWatch and Bertelsmann Stiftung sheds light on what role automated decision-making (ADM) systems play in our lives. As a result of the most comprehensive research on the issue conducted in Europe so far, the report covers the current use of and policy debates around ADM systems in 16 European countries and at EU level.}, keywords = {ai, automated decision making, frontpage, Technologie en recht}, }

Discrimination, artificial intelligence, and algorithmic decision-making external link

vol. 2019, 2019

Abstract

This report, written for the Anti-discrimination department of the Council of Europe, concerns discrimination caused by algorithmic decision-making and other types of artificial intelligence (AI). AI advances important goals, such as efficiency, health and economic growth but it can also have discriminatory effects, for instance when AI systems learn from biased human decisions. In the public and the private sector, organisations can take AI-driven decisions with farreaching effects for people. Public sector bodies can use AI for predictive policing for example, or for making decisions on eligibility for pension payments, housing assistance or unemployment benefits. In the private sector, AI can be used to select job applicants, and banks can use AI to decide whether to grant individual consumers credit and set interest rates for them. Moreover, many small decisions, taken together, can have large effects. By way of illustration, AI-driven price discrimination could lead to certain groups in society consistently paying more. The most relevant legal tools to mitigate the risks of AI-driven discrimination are nondiscrimination law and data protection law. If effectively enforced, both these legal tools could help to fight illegal discrimination. Council of Europe member States, human rights monitoring bodies, such as the European Commission against Racism and Intolerance, and Equality Bodies should aim for better enforcement of current nondiscrimination norms. But AI also opens the way for new types of unfair differentiation (some might say discrimination) that escape current laws. Most non-discrimination statutes apply only to discrimination on the basis of protected characteristics, such as skin colour. Such statutes do not apply if an AI system invents new classes, which do not correlate with protected characteristics, to differentiate between people. Such differentiation could still be unfair, however, for instance when it reinforces social inequality. We probably need additional regulation to protect fairness and human rights in the area of AI. But regulating AI in general is not the right approach, as the use of AI systems is too varied for one set of rules. In different sectors, different values are at stake, and different problems arise. Therefore, sector-specific rules should be considered. More research and debate are needed.

ai, discriminatie, frontpage, kunstmatige intelligentie, Mensenrechten

Bibtex

Report{Borgesius2019, title = {Discrimination, artificial intelligence, and algorithmic decision-making}, author = {Zuiderveen Borgesius, F.}, url = {https://rm.coe.int/discrimination-artificial-intelligence-and-algorithmic-decision-making/1680925d73}, year = {0208}, date = {2019-02-08}, volume = {2019}, pages = {}, abstract = {This report, written for the Anti-discrimination department of the Council of Europe, concerns discrimination caused by algorithmic decision-making and other types of artificial intelligence (AI). AI advances important goals, such as efficiency, health and economic growth but it can also have discriminatory effects, for instance when AI systems learn from biased human decisions. In the public and the private sector, organisations can take AI-driven decisions with farreaching effects for people. Public sector bodies can use AI for predictive policing for example, or for making decisions on eligibility for pension payments, housing assistance or unemployment benefits. In the private sector, AI can be used to select job applicants, and banks can use AI to decide whether to grant individual consumers credit and set interest rates for them. Moreover, many small decisions, taken together, can have large effects. By way of illustration, AI-driven price discrimination could lead to certain groups in society consistently paying more. The most relevant legal tools to mitigate the risks of AI-driven discrimination are nondiscrimination law and data protection law. If effectively enforced, both these legal tools could help to fight illegal discrimination. Council of Europe member States, human rights monitoring bodies, such as the European Commission against Racism and Intolerance, and Equality Bodies should aim for better enforcement of current nondiscrimination norms. But AI also opens the way for new types of unfair differentiation (some might say discrimination) that escape current laws. Most non-discrimination statutes apply only to discrimination on the basis of protected characteristics, such as skin colour. Such statutes do not apply if an AI system invents new classes, which do not correlate with protected characteristics, to differentiate between people. Such differentiation could still be unfair, however, for instance when it reinforces social inequality. We probably need additional regulation to protect fairness and human rights in the area of AI. But regulating AI in general is not the right approach, as the use of AI systems is too varied for one set of rules. In different sectors, different values are at stake, and different problems arise. Therefore, sector-specific rules should be considered. More research and debate are needed.}, keywords = {ai, discriminatie, frontpage, kunstmatige intelligentie, Mensenrechten}, }

Automated Decision-Making Fairness in an AI-driven World: Public Perceptions, Hopes and Concerns external link

Araujo, T., Vreese, C.H. de, Helberger, N., Kruikemeier, S., Weert, J. van,, Bol, N., Oberski, D., Pechenizkiy, M., Schaap, G. & Taylor, L.
2018

Abstract

Ongoing advances in artificial intelligence (AI) are increasingly part of scientific efforts as well as the public debate and the media agenda, raising hopes and concerns about the impact of automated decision making across different sectors of our society. This topic is receiving increasing attention at both national and cross- national levels. The present report contributes to informing this public debate, providing the results of a survey with 958 participants recruited from high-quality sample of the Dutch population. It provides an overview of public knowledge, perceptions, hopes and concerns about the adoption of AI and ADM across different societal sectors in the Netherlands. This report is part of a research collaboration between the Universities of Amsterdam, Tilburg, Radboud, Utrecht and Eindhoven (TU/e) on automated decision making, and forms input to the groups’ research on fairness in automated decision making.

ai, algoritmes, Artificial intelligence, automated decision making, frontpage

Bibtex

Report{Araujo2018, title = {Automated Decision-Making Fairness in an AI-driven World: Public Perceptions, Hopes and Concerns}, author = {Araujo, T. and Vreese, C.H. de and Helberger, N. and Kruikemeier, S. and Weert, J. van, and Bol, N. and Oberski, D. and Pechenizkiy, M. and Schaap, G. and Taylor, L.}, url = {https://www.ivir.nl/publicaties/download/Automated_Decision_Making_Fairness.pdf}, year = {1005}, date = {2018-10-05}, abstract = {Ongoing advances in artificial intelligence (AI) are increasingly part of scientific efforts as well as the public debate and the media agenda, raising hopes and concerns about the impact of automated decision making across different sectors of our society. This topic is receiving increasing attention at both national and cross- national levels. The present report contributes to informing this public debate, providing the results of a survey with 958 participants recruited from high-quality sample of the Dutch population. It provides an overview of public knowledge, perceptions, hopes and concerns about the adoption of AI and ADM across different societal sectors in the Netherlands. This report is part of a research collaboration between the Universities of Amsterdam, Tilburg, Radboud, Utrecht and Eindhoven (TU/e) on automated decision making, and forms input to the groups’ research on fairness in automated decision making.}, keywords = {ai, algoritmes, Artificial intelligence, automated decision making, frontpage}, }

Before the Singularity: Copyright and the Challenges of Artificial Intelligence external link

González Otero, B., & Quintais, J.
Kluwer Copyright Blog, vol. 2018, 2018

ai, Copyright, frontpage

Bibtex

Article{Otero2018, title = {Before the Singularity: Copyright and the Challenges of Artificial Intelligence}, author = {González Otero, B., and Quintais, J.}, url = {http://copyrightblog.kluweriplaw.com/2018/09/25/singularity-copyright-challenges-artificial-intelligence/}, year = {0927}, date = {2018-09-27}, journal = {Kluwer Copyright Blog}, volume = {2018}, pages = {}, keywords = {ai, Copyright, frontpage}, }