The Court of Justice of the European Union (CJEU) has finally delivered its long-awaited judgment in Mio/konektra, C-580/23 and C-795/23, concerning three key issues: (1) copyright treatment of works of applied art; (2) proof of copyright subsistence; and (3) proof of copyright infringement.
(1) Whether originality is the only requirement (also) for works of applied art
The first issue for the CJEU to address concerned the requirements for copyright protection of works of applied art. No surprises here: yes, originality is the only requirement.
The CJEU (quite elegantly) intended this question as entailing a clarification of Cofemel [IPKat here]. However, things date back to much earlier than that (2019): it’s been clear since Flos (2011) [IPKat here] that works of applied art must be treated like any other type of work (according to the Court, at least pre-design reform).
What is then the impact of all of this first part of the judgment? It’s at least a formal one: Member States that have statutory requirements that go beyond originality to protect works of applied art under copyright will now – at last! – need to remove such language.
To exemplify: Italy will need to remove its “artistic value” requirement from the statute. Such a change should have occurred years ago, but at least now it could not be clearer that, yes, that is what is required.
The potential downside? That Member States change the language, but courts keep asking for the same things, now to establish 'originality' rather than 'artistic value'. This change might, therefore, prove more cosmetic than substantive – a ‘Gattopardo-esque’ shift, as Tancredi famously put it: "If we want things to stay as they are, things will have to change."
(2) How to prove copyright subsistence
Next, the Court turned to the thorny question of proving copyright subsistence: what does originality really mean? Like the meaning of life, originality resists a single definition. The Court’s answer? It depends.
The referring courts had listed a number of elements that could be taken into account: the author’s intentions, the existence of alternatives, derivation, post-event circumstances, etc. Like the AG, the Court considered that their value and relevance depend on the circumstances at issue. In a nutshell, anything that could prove “the expression of free and creative choices reflecting the personality of its author” is good enough to be taken into account. In any event, originality does not mean prettiness: “the fact that a design generates, beyond its practical purpose, a specific and aesthetically or artistically significant visual effect is not, in itself, such as to justify that design being classified as a ‘work’” (para 68).
Correctly (and, again, unsurprisingly), the Court noted that, if you wish to sue someone for copyright infringement, you must first prove why your work is protected in the first place: “the creative nature of the choices made by the author of the subject matter cannot be presumed.” (para 65)
Another key (and correct) aspect worth flagging is that the CJEU dismissed the relevance of the author’s intentions as such: “the author’s intentions lie within the realm of ideas. They can therefore be protected only in so far as the author has expressed them in the work concerned.” This is a welcome clarification. To exemplify, it is irrelevant if one was under the effect of substances or fully incapacitated as opposed to perceiving themselves as being a great artist when creating a work: what matters is solely whether the work reflects their personality through the making of free and creative choices.
(3) How to prove copyright infringement
Finally, the Court addressed infringement: should the Infopaq originality-based test [IPKat here] still apply, or is a new approach needed? The AG suggested the latter, by proposing a recognizability-based approach borrowed from (the mess that is) Pelham I [IPKat here] relating to the phonogram producer’s related right of reproduction.
The CJEU adopted some nuances compared to the AG Opinion. While reaffirming the originality-based test, the Court added a 'safeguard': copied original choices must also be recognizable in the infringing work (para 86).
Now, I used the brackets above because I am not sure how this approach could work or be better than the Infopaq test. Like in Pelham I, the CJEU did NOT say according to whom such a recognizability should subsist.
If we take for example the works at issue in the background proceedings to Mio and konektra, recognizability would work very differently for, say, an interior designer as opposed to a lay person or even an AI model trained on hundreds of thousands of furniture designs from different ages and regions of the world. If we move things to the realm of AI and AI-generated outputs, things do not look brighter. This recognizability-based approach, in my view, will create endless chaos for litigants and courts alike.
Conclusion
In short, if you liked AG Szpunar’s Opinion, you’ll like the Mio/konektra judgment too. But if you hoped for definitive clarity after 15+ years since Infopaq, disappointment awaits. EU copyright law, like love in The Leopard, burns brightly for a year – then leaves ashes for decades. Till the next referral then …
[Originally published on The IPKat on 5 December 2025]