The upcoming Mio/konektra judgment: What the CJEU should decide regarding the copyright infringement test

Back in the spring, The IPKat reported on the then fresh Opinion of Advocate General (AG) Szpunar in the Mio/konektra joined referrals.

The questions referred by the Svea Court of Appeal, Patent and Commercial Court of Appeal (Sweden) in Mio, C-580/23 and the Federal Court of Justice (Germany) in konektra, C-795/23 go to the very heart of copyright, tackling key questions like: (1) when does copyright arise and how is originality to be assessed?, and (2) when is copyright infringed and what is the test to be adopted to this end?

When the Court of Justice of the European Union (CJEU) issues its judgment in the joined cases, we will likely have yet another crucial building block in the EU copyright architecture.

In this post, I will revisit the AG’s approach to the second issue above – that is: the infringement test – and try to articulate why the CJEU should NOT ‘depart’ from the approach established in yet another seminal decision (Infopaq). The Court should confirm that the copyright infringement test is based on originality, not – as proposed by the AG – the recognizability of what has been copied.

Why the copyright infringement test is based on originality and should continue being so

In Infopaq, the CJEU established that there is reproduction “in part” under Article 2(a) of the InfoSoc Directive when the part of a work that has been copied is original in the sense that it is its author’s own intellectual creation. The Court, which has long regarded such an approach as settled, has progressively refined the concept of originality.

Originality is found when – through the choice, sequence and combination of elements like words, figures or mathematical concepts – an author expresses their creativity in an original manner and exercises “creative freedom” (FAPL, [98]). In all of this “mere intellectual effort and skill” in creating a work are not relevant to establishing originality (Funke Medien, [23]), nor is when the creation of a work “is dictated by technical considerations, rules or constraints which leave no room for creative freedom” (ex multis, Cofemel, [31]).

In sum: there is reproduction ‘in part’ of a work when the defendant has copied an original part thereof. It is obvious that – to determine actual infringement – consideration of other aspects is required, including the applicability of any available exception or limitation.

While it might be argued that what makes a work ‘original’ depends on the circumstances at hand and that there might not be any hard and fast rule (which AG Szpunar also correctly acknowledged in his Mio/konektra Opinion), a test based on originality is not only appropriate, but is the only right one.

That is so for several reasons, including the basic consideration that, as the CJEU has consistently held over 15+ years, originality is the sole requirement that needs to be fulfilled for copyright to vest in a work: nothing more, nothing less.

In turn, not every copying is actionable: only copying of an original part is. In this sense, originality is like a coin with two faces: ‘heads’ is when copyright arises, ‘tails’ is when copyright is prima facie infringed.

It is unclear what advantages the adoption of a recognizability approach – as opposed to originality – would offer: if anything it could lend its side to greater arbitrariness and subjectivity, the latter being something that AG Szpunar himself warned against.

One key shortcoming is that neither the CJEU in Pelham I nor AG Szpunar in Mio/konektra indicated the benchmark for assessing recognizability: should it be so to an expert (note: unlike other IP rights, copyright does not envisage any fictional ‘character’ like, e.g., the person skilled in the art, the informed user or even the average consumer), to an average person, or be even machine-detectable? Depending on which one is chosen, the border between lawful and unlawful copying would shift greatly

An objective, not subjective test

Indeed, the infringement test must be objective, in line with international and EU law. What matters is:

1. If there has been derivation. By ‘derivation’ it is meant direct or indirect access to the earlier copyright work, which can be proved in different ways, as well as by means of suitable inferences indicating that access to the work was more likely to occur than not (e.g., due to the competitor/claimant holding a prominent position in the market and/or having displayed and/or advertised their work, etc); if so
2. Whether there has been copying of protected features decided by an objective comparison of the earlier copyright work and the later work; if so
3. If any defence applies in line with the closed optional catalogue under Art 5 of the InfoSoc Directive and applicable national law.

International and EU copyright legislation is clear that reproduction is to be approached objectively and there is no indication to the contrary anywhere else. Consideration of subjective elements might be relevant to damage determination (in line with the de minimis guidance provided by the Enforcement Directive), not the prior establishment of liability.

With the exception of situations like linking and platform liability, which the CJEU brought under the umbrella of the harmonized EU right of communication/making available to the public – preferring that to leaving these matters to unharmonized national laws: see Opinion of AG Szpunar in Ziggo, [3]), the Court has consistently and actively rejected assessments based on subjective elements, including having regard to works of applied art.

The Court itself has reiterated on multiple occasions that no element of subjectivity should be present in the process of identifying the protected subject-matter (see, e.g., Levola Hengelo, [40], and Cofemel, [32]). In line with the wording of the international and EU right of reproduction, the same goes for the prima facie test to determine direct/primary liability for unauthorized reproductions. In this sense, it may be recalled that past suggestions of AGs (as it was for example the case of AG Campos in his Opinion in Brompton) to consider the infringer’s intentions were subsequently dismissed by the CJEU (and national courts too). Also, in Mio/konektra AG Szpunar himself rejected consideration of the author’s intention.

In sum: the correct approach to establishing liability for primary/direct copyright infringement is the 3-step process indicated above: derivation – actionable copying (based on originality) – lack of exceptions or limitations (or other defences).

Reproduction entails derivation

There is another aspect of the AG Opinion that is ambiguous and thus potentially problematic: at [58], [72], and [74], the Opinion discusses what happens when two authors have made very similar, or even identical, choices.

On the one hand, it is correctly suggested that such a situation does not rule out the possibility of two self-standing copyrights, subject to those choices being ‘free and creative’ and thus original. On the other hand, the Opinion holds that “the mere possibility of such an independent creation cannot however justify a refusal of copyright protection where the reproduction of creative elements of the protected work has been established.”

Such a statement needs to be clarified: it is true that two authors may come up with similar works that are both protected and neither is infringing of the other’s copyright, but if there is 'reproduction' (as the AG refers to it) of protected elements, that is clearly derivation and therefore a prima facie infringement of the first author’s copyright. Above all, it is unclear why someone else’s subsequent ‘reproduction’ of protected elements should be at all relevant to – even determinative of! – whether copyright subsist in the first place in one’s work.

Holding otherwise would be contradictory and also run contrary to basic tenets of copyright law, which have been long endorsed by the CJEU itself. One might also consider such an approach, if the CJEU were to endorse it, one that unduly introduces a new copyright exception or limitation beyond the closed catalogue of Article 5 of the InfoSoc Directive, this being a possibility – as the CJEU also adamantly held in 2019 in Funke Medien and Spiegel Online and, in late 2024, in Kwantum – that is precluded under EU law.

While it is true that some national courts in the past have regarded reproduction with elaboration (that is: the addition of one’s own originality) something that would rule out infringement in the first place, this is not how the CJEU has approached the prima facie infringement test. Again, the Court has consistently endorsed the 3-step approach indicated above. More recently, attentive national case law has indicated the need to comply with the CJEU-mandated approach and thus 'shelf' older national cases decided prior to the CJEU’s harmonizing stance on infringement.

Conclusion

Mio/konektra promises to be one of the most relevant CJEU copyright judgments ever issued to date. The Court has an opportunity to reflect on and put a full stop (or, at least, a semi-colon!) on an evolution, driven by the Court itself, which has been unfolding since the 2009 Infopaq judgment.

While a lot of the AG Opinion should be endorsed in the resulting ruling, the infringement test should remain one based on originality and one where derivation (whether direct or indirect) plus reproduction of an original part still equates prima facie infringement.

[Originally published on The IPKat on 18 September 2025]