Suchergebnisse

Abstract
This paper focuses on the two exceptions for text and data mining (TDM) introduced in the Directive on Copyright in the Digital Single Market (CDSM). While both are mandatory for Member States, Art. 3 is also imperative and finds application in cases of text and data mining for the purpose of scientific research by research and cultural institutions; Art. 4, on the other hand, permits text and data mining by anyone but with rightholders able to 'contract-out' (Art. 4). We trace the context of using the lever of copyright law to enable emerging technologies such as AI and the support innovation. Within the EU copyright intervention, elements that may underpin a transparent legal framework for AI are identified, such as the possibility of retention of permanent copies for further verification. On the other hand, we identify several pitfalls, including an excessively broad definition of TDM which makes the entire field of data-driven AI development dependent on an exception. We analyse the implications of limiting the scope of the exceptions to the right of reproduction; we argue that the limitation of Art. 3 to certain beneficiaries remains problematic; and that the requirement of lawful access is difficult to operationalize. In conclusion, we argue that there should be no need for a TDM exception for the act of extracting informational value from protected works. The EU's CDSM provisions paradoxically may favour the development of biased AI systems due to price and accessibility conditions for training data that offer the wrong incentives. To avoid licensing, it may be economically attractive for EU-based developers to train their algorithms on older, less accurate, biased data, or import AI models already trained abroad on unverifiable data.

What factors lead a copyright owner to request removal of potentially infringing user-generated content? So-called "notice-and-takedown" measures are provided in the United States under Section 512 of the U.S. Copyright Act (as amended by the Digital Millennium Copyright Act 1998) and enabled in the European Union under the Directive on Electronic Commerce (2000/31/EC). While the combination of limiting liability ("safe harbor") and notice-and-takedown procedures was originally conceived as a means of balancing innovation with the interests of rightholders, there has been limited empirical study regarding their effects. This research investigates, for the first time, the factors that motivate takedown of user-generated content by copyright owners. We study takedowns within an original dataset of 1,839 YouTube music video parodies observed between January 2012 and December 2016. We find an overall rate of takedowns within the sample of 32.9% across the 4-year period. We use a Cox proportional hazards model to investigate propositions from rightholder groups about the factors that motivate takedowns: these include concerns about commercial substitution; artistic/moral concerns; cultural differences between firms; and YouTube uploader practices. The main finding is that policy concerns frequently raised by rightholders are not associated with statistically significant patterns of action. For example, the potential for reputational harm from parodic use does not appear to predict takedown behavior. Nor does commercial popularity of the original music track trigger a systematic response from rightholders. Instead, music genre and production values emerge as significant factors. We suggest that evolving policy on intermediary liability - for example with respect to imposing filtering systems (automatically ensuring "stay-down" of potentially infringing content) - should be carefully evaluated against evidence of actual behavior, which this study shows may differ materially from stated policy positions.

What factors lead a copyright owner to request removal of potentially infringing user-generated content? So-called "notice-and-takedown" measures are provided in the United States under Section 512 of the U.S. Copyright Act (as amended by the Digital Millennium Copyright Act 1998) and enabled in the European Union under the Directive on Electronic Commerce (2000/31/EC). While the combination of limiting liability ("safe harbor") and notice-and-takedown procedures was originally conceived as a means of balancing innovation with the interests of rightholders, there has been limited empirical study regarding their effects. This research investigates, for the first time, the factors that motivate takedown of user-generated content by copyright owners. We study takedowns within an original dataset of 1,839 YouTube music video parodies observed between January 2012 and December 2016. We find an overall rate of takedowns within the sample of 32.9% across the 4-year period. We use a Cox proportional hazards model to investigate propositions from rightholder groups about the factors that motivate takedowns: these include concerns about commercial substitution; artistic/moral concerns; cultural differences between firms; and YouTube uploader practices. The main finding is that policy concerns frequently raised by rightholders are not associated with statistically significant patterns of action. For example, the potential for reputational harm from parodic use does not appear to predict takedown behavior. Nor does commercial popularity of the original music track trigger a systematic response from rightholders. Instead, music genre and production values emerge as significant factors. We suggest that evolving policy on intermediary liability - for example with respect to imposing filtering systems (automatically ensuring "stay-down" of potentially infringing content) - should be carefully evaluated against evidence of actual behavior, which this study shows may differ materially from stated policy positions.

This document presents an edited transcript of the one-day event, 'Research Perspectives on the Public Domain', held at the University of Glasgow on 11thOctober, 2013. The public domain is a subject of vital interest to legal scholars, but its implications are far reaching – indeed, the public domain concept is germane to subjects as diverse as film and media studies, economics, political science and organisational theory. It was a central purpose of the workshop to arrive at a workable definition of the public domain suitable for empirical investigation. The traditional definition (1) takes the copyright term as the starting point, and defines the public domain as "out of copyright", i.e. all uses of a copyright work are possible. A second, more fine-grained definition (2) still relies on the statutory provisions of copyright law, and asks what activities are possible with respect to a copyright work without asking for permission (e.g. because use is related to "underlying ideas" not appropriating substantial expressions, or because use is covered by specific copyright exceptions). A third definition (3) includes as part of the public domain all uses that are possible under permissive private ordering schemes (such as creative commons licences). A forth definition (4) moves into a space that includes use that would formally be copyright infringement but is endorsed, or at least tolerated by certain communities of practice (e.g. machinima or fan fiction). The conference was designed to test these definitional approaches, and national and international speakers from relevant disciplinary fields were invited to share their research projects, with a particular focus on the underlying concept of the public domain. This document is a citable documentation of those presentations, along with a panel discussion that followed. This event was funded through a Knowledge Exchange grant, 'Valuing the Public Domain', from the Economic and Social Research Council (ESRC ES/K008137/1) and the UK Intellectual Property Office ...

"What can and can't be copied is a matter of law, but also of aesthetics, culture, and economics. The act of copying, and the creation and transaction of rights relating to it, evokes fundamental notions of communication and censorship, of authorship and ownership--of privilege and property. This volume conceives a new history of copyright law as fifteen leading academics discuss the changing state of intellectual property across time and between countries"--Publisher's description

Intro -- Privilege and Property -- Contents -- Contributors -- Introduction. -- 1. From Gunpowder to Print: The Common Origins of Copyright and Patent -- 2. 'A Mongrel of Early Modern Copyright': Scotland in European Perspective -- 3. The Public Sphere and the Emergence of Copyright: Areopagitica, the Stationers' Company, and the Statute of Anne -- 4. Early American Printing Privileges. The Ambivalent Origins of Authors' Copyright in America -- 5. Author and Work in the French Print Privileges System: Some Milestones -- 6. A Venetian Experiment on Perpetual Copyright -- 7. Les formalités sont mortes, vive les formalités! Copyright Formalities and the Reasons for their Decline in Nineteenth Century Europe -- 8. The Berlin Publisher Friedrich Nicolai and the Reprinting Sections of the Prussian Statute Book of 1794 -- 9. Nineteenth Century Controversies Relating to the Protection of Artistic Property in France -- 10. Maps, Views and Ornament: Visualising Property in Art and Law. The Case of Pre-modern France -- 11. Breaking the Mould? The Radical Nature of the Fine Arts Copyright Bill 1862 -- 12. 'Neither Bolt nor Chain, Iron Safe nor Private Watchman, Can Prevent the Theft of Words': The Birth of the Performing Right in Britain -- 13. The Return of the Commons - Copyright History as a Common Source -- 14. The Significance of Copyright History for Publishing History and Historians -- 15. Metaphors of Intellectual Property -- Bibliography -- Index.

AbstractMachine learning, a subfield of artificial intelligence (AI), relies on large corpora of data as input for learning algorithms, resulting in trained models that can perform a variety of tasks. While data or information are not subject matter within copyright law, almost all materials used to construct corpora for machine learning are protected by copyright law: texts, images, videos, and so on. There are global policy moves to address the copyright implications of machine learning, in particular in the context of so-called "foundation models" that underpin generative AI. This paper takes a step back, exploring empirically three technological settings through detailed case studies. We set out the established industry methodology of a lifecycle of AI (collecting data, organising data, model training, model operation) to arrive at descriptions suitable for legal analysis. This will allow an assessment of the challenges for a harmonisation of rights, exceptions and disclosure under EU copyright law. The three case studies are:



Machine learning for scientific purposes, in the context of a study of regional short-term letting markets;





Natural Language Processing (NLP), in the context of large language models;





Computer vision, in the context of content moderation of images.


We find that the nature and quality of data corpora at the input stage is central to the lifecycle of machine learning. Because of the uncertain legal status of data collection and processing, combined with the competitive advantage gained by firms not disclosing technological advances, the inputs of the models deployed are often unknown. Moreover, the "lawful access" requirement of the EU exception for text and data mining may turn the exception into a decision by rightholders to allow machine learning in the context of their decision to allow access. We assess policy interventions at EU level, seeking to clarify the legal status of input data via copyright exceptions, opt-outs or the forced disclosure of copyright materials. We find that the likely result is a fully copyright-licensed environment of machine learning that may have problematic effects for the structure of industry, innovation and scientific research.