The impact of AI on European creative sectors has been a key topic for EU policymakers over the past year, and the Danish 2025 EU Presidency has placed it high on its agenda. In line with this focus, on 6th October 2025, AEPO-ARTIS (Association of European Performers’ Organisations), together with Danish organisations FILMEX and DMF (Dansk Musiker Forbund), hosted “The Impact of Artificial Intelligence on the Jobs of Creative Workers,” an event addressing the protection of performers’ voice, likeness, and artistic contributions from unauthorised AI use.
No better location than the Workers Museum in Copenhagen to bring together representatives of collective management organisations from across Europe with Danish performers, unions and policymakers, and provide its ongoing exhibition on how industrial history has changed the course and conditions of workers’ lives with a one day extension focussed on our actors and musicians.
Tilo Gerlach, President of AEPO-ARTIS, opened the meeting highlighting performers’ central role in AI development:
“Generative AI is based on the contribution of performers and their creativity. It is not more than fair that they benefit from any income generated by AI. Full transparency of the use of their Performances is needed as well as an appropriate remuneration.”
Thomas Sandberg, renowned performer and Chair of the Danish Musicians’ Union, followed by putting focus on the role of collective action and the strong reputation Denmark has been able to build up in this regard. The opening speeches were concluded by Peter Skaarup, Member of the Danish Parliament, who addressed the ongoing Danish proposal on protecting performers against AI-generated content.
The first panel, “Voice and likeness theft by AI,” moderated by Nick Yule, AEPO-ARTIS Head of Legal, featured Maria Fredenslund, CEO and founder of the Danish Rights Alliance, who stressed the need for transparency:
“’Opt-out’ is a shadow argument used by AI developers to reason their extensive use of content to train AI machines like GPT and Llama. Just like the argument that online services such as YouTube used years back, when they claimed that they were free from liability for activities on their platform because they were just neutral hosting providers. It took 15 years to correct that mess. Let’s not repeat that. Real and meaningful transparency on training input is necessary, as well as clear regulation stating that output made with actors’ voices and musicians’ music is illegal and must not be distributed online.”
AI’s threats to audiovisual performers were highlighted by Patricia Riera Barsallo from Spanish collective management organization AISGE :
“The legal framework should grant AV performers the right to authorize the use of their performances in the training of AI systems in exchange for fair remuneration. The right should be unwaivable and subject to mandatory collective management to make it effective. Also, it should grant a mechanism to ensure that AV performers receive fair compensation for the exploitation of outputs generated by AI.”
The challenges of distinguishing real from manipulated content online were addressed by Zebbie Gaya, from the Danish Ministry of Culture, who noted that there are increasingly difficulties of being able to distinguish between between authentic and manipulated material online which can, for example when manipulating videos create fundamental doubt about what is a true representation of reality, highlighting the need for regulation:
“The draft bill introducing amendments to the Danish Copyright Act on digital imitations of personal characteristics etc. will hopefully be presented for the parliament this winter. It will be interesting to follow the area of AI and copyright also years from now.” she said.
In a second panel focus was shifted to “Making AI work for musicians.” Danish musician Troel Abrahamson, who also provided a musical intermezzo, emphasized the need to regulate the way AI is used, rather than technology itself stating:
“As someone who works with AI both artistically and professionally I am concerned that we risk throwing out the baby out with bathwater by regulating technology and not use. There’s a lot of interesting and exciting work being done on the grassroots level, and we need to recognize that, support it and not inadvertently stifle it. That requires careful consideration and conceptualization.“
Thomas Sandberg reinforced the need for legal safeguards:
“We need transparency according to the training of AI engines, fair licensing models and remuneration accompanied by copyright legislation that will protect the expression of the artform and artistic identity.”
Marion Argiolas, from French collective management organisation SPEDIDAM, highlighted musicians’ vulnerability in the AI ecosystem:
“At the moment I don’t really see AI working for musicians. It is rather the contrary: musicians are working for A. And they are doing it for free.”
The event emphasised the urgent need for EU and national frameworks that ensure transparency and fair remuneration to protect performers’ rights in the era of AI and the importance of collective management as a crucial partner in make this work in practice.
