Reservation for text and data mining for use in artificial intelligence
Stim wants to ensure that the copyright-protected works represented by Stim are not in any way incorporated into the development and training of AI without a license permitting this. This includes, but is not limited to, language models, foundation models, generative AI systems, recommendation algorithms and the collection of metadata/statistics.
Stim is a collective management organisation that, under applicable copyright law, promotes and administers the economic rights to music belonging to the natural or legal persons who are members of or are affiliated with Stim (hereinafter jointly referred to as "Rights Holders").
Stim expressly reserves the right of reproduction for text and data mining in accordance with Chapter 2, Section 15a of the Copyright Act (1960:729) on copyright in literary and artistic works (the "Copyright Act") and Article 4 of the European Parliament and Council Directive (EU) 2019/790 of April 17th on copyright and related rights in the digital single market and amending Directives 96/9/EC and 2001/29/EC.
The reservation means that any use for text and data mining of the works represented by Stim at any given time requires Stim's express permission, unless otherwise provided by law.
Background
The development of AI, including generative AI, is entirely dependent on large volumes of data and its processing. These datasets enable AI systems to identify, learn to understand, and analyze complex patterns, among other things. To facilitate the "learning/training" of an AI effectively, access to extensive and varied datasets is required, including copyright-protected content that Stim administers on behalf of the Rights Holders.
Stim emphasises that the association's view is that Chapter 2, Section 15a of the Copyright Act and the underlying EU legislation do not permit text and data mining for the purpose of developing generative AI systems (which store, process, distribute, or otherwise utilise protected works). Stim is, however, aware that this interpretation of the legal situation is not shared by everyone, particularly not by international tech companies and other actors behind the development and training of AI systems/tools, language models or the like. Regardless of the interpretation made, the relevant rights are reserved by Stim.