The Association of European Performers’ Organisations is addressing how AI companies use recordings of musicians and dancers for machine training without direct performer consent.
Key facts
- •AEPO-ARTIS represents performers across Europe, including musicians and those in the audiovisual sector.
- •Neighbouring rights are defined as unalienable protections that prevent the harmful presentation or distribution of a performer's work.
- •The organization is working to clarify how these rights apply as AI companies increasingly use existing recordings for machine training.
- •Performers often lack direct control over licensing agreements because rights are frequently managed by a chain of record labels and distributors.
- •Ibán García del Blanco serves as the secretary general of AEPO-ARTIS.
The Association of European Performers’ Organisations (AEPO-ARTIS) is currently navigating the legal challenges posed by artificial intelligence companies using recorded performances to train machine learning models. As AI systems increasingly ingest vast amounts of creative work, the organization is highlighting the vulnerability of performers whose contributions are often licensed to AI developers by third-party distributors or record labels without the artists' direct involvement.
The Role of Neighbouring Rights
While authors hold primary copyright, performers—such as orchestral musicians, actors, and dancers—possess 'neighbouring rights.' AEPO-ARTIS secretary general Ibán García del Blanco emphasizes that these rights are unalienable and serve to protect the performer from having their work presented or distributed in ways that could damage their professional reputation. These rights are distinct from authorship and are intended to safeguard the contributions of those who perform, rather than create, cultural works.
Challenges in the Licensing Chain
The current structure of the music and audiovisual industries complicates the protection of performers' rights in the AI era. Often, a performance is recorded by a label, owned by a distributor, and then licensed to an AI company for training purposes. This multi-layered chain of ownership frequently leaves the individual performer, such as an orchestral oboe player, without a direct say in how their performance is utilized by AI developers to train systems capable of regurgitating their work.
Advertisement
This article was independently rewritten by ManyPress editorial AI from reporting originally published by EUobserver, EUobserver.
