Chetanay Wahi
Research Assistant
Media Governance & Industries Research Lab
University of Vienna
In June 2026, the Lab submitted to the European Commission’s Call for Evidence on a better copyright environment for European creativity and innovation (Ares(2026)4845636). The submission was made on behalf of the Horizon Europe project ANIMA MUNDI (ID #101178027) and rests on the project’s first empirical deliverable deriving from five stakeholder workshops, eleven focus groups, and 43 animation-industry participants gathered across Europe between June and September 2025 (D4.1, Sarikakis et al., 2025).
It is worth being precise about what this act is, and what it is not. A call for evidence is an early stage of the EU’s regulatory cycle. It is the point at which a problem is still being defined, before options are weighed in a formal impact assessment. Submissions do not write law. They enter an evidence base that the Commission compiles and summarises for the Parliament and the Council. Whether any single contribution is taken up is uncertain and outside its author’s control. What the stage does offer is leverage over framing; the categories and distinctions fixed now shape which questions the later assessment treats as legitimate. That is where a research lab like ours can be useful. We did not proceed by advocating an outcome, but by insisting that the evidence base contains a category that is otherwise easy to overlook.
Our submission makes that category explicit: the next generation of European creators.
The animation workforce is freelance-dependent and organised around micro-enterprises and long project cycles. Entry-level pipelines produce senior practitioners; intellectual property accumulates at the end of careers but originates at their beginning. This structure places early-career creators at the intersection of the two issues the consultation treats separately. First, the pressure of generative AI on creative labour, and second, the operation of Articles 18–23 of the 2019 Copyright Directive on author remuneration. They are at once the cohort most exposed to AI substitution and the cohort with the weakest position when signing the contracts that decide whether remuneration, transparency, and revocation rights ever take effect. Assessed in isolation, each issue looks manageable. Assessed together, they compound at a single point in the career, and that point is invisible in most stakeholder consultations, which are populated by established professionals and their representatives.
Two claims follow from this framing.
The first concerns AI. The consultation treats “use of works by AI” as a single phenomenon. It is not. Training a model on protected works and using protected works at inference are distinct acts with different legal statuses and different economics. Training raises an attribution problem that resists individual remedy and points toward collective or levy-based instruments; inference is in principle attributable and can support licensing. Collapsing the two produces an instrument that under-addresses both. For creators without an organised rights-holding capacity (which describes most people at the start of a career) the distinction is not academic. It determines whether any remedy exists at all. We argue the distinction should be treated as an analytical prior in the impact assessment, established before policy options are considered rather than after.
The second concerns remuneration. Articles 18–23 were built on an explicit recognition that authors negotiate from a structurally weak position. But the machinery meant to correct for that weakness (transparency reporting, contract adjustment, revocation, and the non-waivability rule that protects them) presupposes that a creator knows the rights exist, holds the information needed to invoke them, and can absorb the cost and reputational risk of doing so in a small industry with relationship-based hiring. Our workshop evidence indicates that none of these conditions reliably holds at the junior end. Buy-out contracts and lump-sum payments have become a sectoral default; national tax-rebate schemes are functioning, in effect, as subsidies that move European IP to non-EU platforms; and the transparency provision meant to be the framework’s cornerstone assumes a line of sight to exploitation that the individual animator, illustrator, or background artist does not have. A protective architecture that the weakest party cannot see or use is, for that party, not protection.
None of this guarantees uptake, and it would be a misrepresentation of the process to imply otherwise. Consultation responses are many; the Commission is under no obligation to adopt any of them; and the causal path from a submission to a clause in a directive is diffuse and rarely traceable. We submit anyway, for two reasons. Holding the evidence within the project and its academic outputs would leave the policy record poorer at exactly the moment framing is set. And because a summary of the feedback reaches the Parliament and Council as the legislative debate proceeds, the submission is a route by which primary evidence (not opinion) is placed on that record while it can still matter. Making findings usable at that moment is, we take it, part of the work of a lab that studies governance than just describing it.
The full submission, including its recommendations for the impact assessment and for the parallel review of the Copyright Directive, is available via the EU consultation portal (link). It will not, on its own, change the law. What it does is put a specific, evidenced claim on the record, that a reform designed to protect European creativity should be tested against its effect on the people who will still be creating when the reform takes hold.
Reference
Keinonen, H., Wahi, C., Sarikakis, K., Rossato Fernandes, M., Tran, T., Tinen, P., Vlassis, A., Goffredo, S., Dâmaso, M., Turan, P., Sganga, C., De Potesta, A., Afilipoaie, A., & Ranaivoson, H. (2025). D4.1 – Set-the-stage report on international promotion of European Animation Industry. Horizon Europe project ANIMA MUNDI (ID: 101178027). https://doi.org/10.5281/zenodo.17482738
This image was generated entirely using Google’s generative AI, powered by the Gemini ‘Pro’ model.
Leave a Reply