The Impact of Copyright and Personal Data Laws on the Creation and Use of Models for Language Technologies

Abstract

The authors address the legal issues relating to the creation and use of language models. The article begins with an explanation of the development of language technologies. The authors analyse the technological process within the framework copyright, related rights and personal data protection law. The authors also cover commercial use of language models. The authors' main argument is that legal restrictions applicable to language data containing copyrighted material and personal data usually do not apply to language models. Language models are generally not considered derivative works. Due to a wide range of language models, this position is not absolute.

EU Data Governance Act: Outlining a Potential Role for CLARIN

“…It is similar to definitions of research and language data and therefore entails similar challenges. For further discussion, see Kelli et al (2020); Kelli et al (2018).…”

The Interaction of Personal Data, Intellectual Property and Freedom of Expression in the Context of Language Research

“…An overarching theme for this and the following section concerns legal obligations relating to derived data, e.g., data derived through text and data mining (TDM). For further discussion, see Kelli et al (2020). Interestingly, the TDM exception contained in the DSM Directive does not require attribution.…”

“…6 bis , which sets the international standard and binds all current EU member states (and almost all of the remaining world), gives authors the attribution (paternity) right. The relevant question here is whether a researcher who has collected language data containing copyrighted content (for further discussion on the process of development of language technologies from the legal perspective, see Kelli et al 2020) should attribute the author of the content or follow the data minimisation principle and remove all personal data (e.g., the author's name) that is not necessary for processing.…”

Community-Based Survey and Oral Archive Infrastructure in the Archivio Vi.Vo. Project

“…The project will also exploit the federated identity service of the CLARIN infrastructure, in order to manage users' access. A robust system for managing authentication is essential for audio and audiovisual archives because of the frequent privacy, ownership, and copyright issues concerning their content Kelli et al, 2020). Several classes of users are taken into consideration, each of them with different access grants.…”