This case concerned the famous Brompton bicycle which can be folded to carry away after use (Case C‑833/18, SI, Brompton Bicycle Ltd. v. Chedech / Get2Get); the author commented on the Advocate General’s opinion here and here. The bike was once protected by a patent and, following its expiry, the defendant (Get2Get) started selling a similar bike, called Chedech, in Belgium (the designs of the two bikes are depicted in the AG Opinion). Brompton then sued for copyright infringement. Get2Get argued that the appearance of the bike is dictated by the technical function, namely that the bicycle can fold into three different positions, therefore it could only be protected by patent law and not copyright law. In Belgium, the multiplicity of shapes theory, according to which a shape is not considered necessary to achieve a technical result if there is proof that other possible shapes are available which allow the same technical result, applies in copyright law. Unsure whether this complied with EU law, Liège’s Companies Court referred two questions to the Court of Justice of the European Union (the Court), which were slightly merged and reworded by the Court as follows:
– Must Articles 2 to 5 of the InfoSoc Directive (2001/29) “be interpreted as meaning that the copyright protection provided for therein applies to a product whose shape is, at least in part, necessary to obtain a technical result” (para. 20, emphasis added).
The Court’s answer is that copyright can protect a product’s shape which is, at least in part, necessary to obtain a technical result, provided it is original (para. 38). If the shape is solely dictated by function, copyright cannot protect it.
Let us analyse the judgment in more detail. The Court starts its analysis by restating its case law on the notion of a work as detailed in its Cofemel v G-Star Raw (itself referring to Infopaq I, Football Dataco v Yahoo! UK and Levola), SAS Institute and BSA rulings. Accordingly, its reasoning goes as follows. For something to be a work, it must be original i.e. the author’s own intellectual creation and be an expression of that creation. This means that “the subject matter reflects the personality of its author, as an expression of his free and creative choices”. These conditions are both necessary and sufficient (Brompton, paras. 22-23). Therefore, “where the realisation of a subject matter has been dictated by technical considerations, rules or other constraints which have left no room for creative freedom”, it cannot be original (para. 24). Last but not least, the subject matter has to be identifiable with sufficient precision and objectivity (para. 25).
The Court then draws the conclusion that “a subject matter satisfying the condition of originality may be eligible for copyright protection, even if its realisation has been dictated by technical considerations, provided that its being so dictated has not prevented the author from reflecting his personality in that subject matter, as an expression of free and creative choices” (para. 26). Therefore, it is only when the subject matter is solely dictated by technical function that copyright cannot subsist in it. The Court supports this conclusion by relying on article 2 of the WIPO Copyright Treaty and its rulings in SAS Institute and BSA, which state that copyright does not protect ideas and when the expression of components of a subject matter “is dictated by their technical function, the different methods of implementing an idea are so limited that the idea and the expression become indissociable” (para. 27).
The national court’s questions do not refer to the requirement that the bike be identifiable with enough precision and objectivity so the Court surmises that the bike passes this test. Thus, the only condition to verify is whether it is original. For the Court, even if the shape of the bike seems dictated by function (the folding mechanism which allows several positions, one of which is that the bike remains stable when folded), originality could still remain. This is something for the national court to decide (paras. 28-30).
The Court then moves to the second part of its analysis, in essence replying to the second question posed by the national court on possible criteria to take into account when assessing whether a shape is necessary to achieve a technical result (para. 35-37). It first specifically states that even if more than one shape can achieve the same technical result, this is not decisive of originality, implicitly rejecting the theory of the multiplicity of shapes the national court was referring to. It also pithily rejects the intention of the infringer as a relevant criterion to decide if the shape is original. As to the last two criteria (the existence of a prior expired patent and the effectiveness of the shape in achieving the same technical result), they “should be taken into account only in so far as those factors make it possible to reveal what was taken into consideration in choosing the shape of the product concerned” (para. 36). Last but not least, the Court ends by stating that the national court must “take account of all the relevant aspects of the present case, as they existed when that subject matter was designed, irrespective of the factors external to and subsequent to the creation of the product” (para. 37). So, national courts must avoid the pitfalls of hindsight.
The Court’s judgment is to be applauded for its correct application of copyright principles and its double rejection of the theory of the multiplicity of shapes and causality theory. Its orthodox and logical, if at times repetitive (paras. 31-34 repeat the main teachings of the judgment stated in paras. 22-27, which may mean the Court simply, and justifiably, wanted to reinforce these points), approach to this thorny issue is to be commended. The reference to its “functionality case law” namely SAS Institute, BSA and Football Dataco v Yahoo! UK in relation to the idea/expression dichotomy, merger doctrine and the originality criterion (paras 24 and 27) gives the Court’s ruling a solid basis and thus strengthens it.
As the Court rejects the theory of the multiplicity of shapes in copyright law, Belgian courts and any other eventual national courts or legislatures will thus have to stop applying it. As to the criterion of the effectiveness of the shape, which refers to the causality approach which the Court adopted in design law (see Doceram), the Court cautiously does not apply it but says it “should be taken into account only in so far as those factors make it possible to reveal what was taken into consideration in choosing the shape of the product concerned” (para. 36, emphasis added). As the European Copyright Society had stated in its opinion on the case, “while a finding that the technical outcome was the only reason for the adoption of that shape may well be conclusive in ruling out protection also in the copyright domain, the reverse does not hold true. Indeed, even though the solution was not exclusively dictated by a technical factor, it remains necessary to identify whether the technical constraints still enabled the designer to express her creative abilities by making free and creative choices in the selection of the shape.” The non-application of the causality approach is to be welcomed: copyright’s aim is not protecting function – in fact functional works (some 3D objects, software and databases) are the odd ones out in copyright law. On the other hand, design law was clearly meant to incentivise functional designs, although not those exclusively dictated by function whose protection is reserved to patent law (art. 8(1) design regulation and 7(1) design directive). As previous CJEU cases themselves show, design law’s goal is to protect the investment incurred in creating a design against imitation by competitors, whereas copyright’s goal is to encourage creativity, among others by reserving the economic exploitation of works to creators and their assignees.
On the other points, the Court adopts in essence the same view as the European Copyright Society’s opinion – namely that the intention of the infringer is irrelevant and that the existence of a prior patent is only indicative and not conclusive of the absence of copyright protection (the AG had also embraced the latter point).
Another positive aspect is that the Court follows none of the AG’s approaches that were in our view ill-judged, namely the application of art. 4(1)(e)(ii) of the trade mark directive (excluding signs consisting exclusively of the shape, or another characteristic, of goods which is necessary to obtain a technical result) by analogy, the application of unfair competition law (specifically slavish imitation) and the use of the designer’s intention. On these, we refer the reader to our post here and to others here and here.
Penumbras and regrets
It is a pity that the Court did not take the opportunity of the Brompton reference to clarify the obscure statement it made in paras. 97-98 of Painer. We refer the reader to our comment on the AG opinion on this point. It was not necessary in order to answer the questions, but it would have been a good thing to do. Less problematic is the Court’s silence on the creator’s intention, which admittedly the national Court did not ask about, but the AG mentioned as an element to take into account. The Court’s total lack of reference to the AG’s opinion (on any point not just that one) probably means that the criterion is irrelevant. Arguably, this is so because of Levola’s rejection of subjective considerations in deciding what a work is, as already mentioned. Indeed, the Court refers to that specific paragraph of Levola (para. 40) in its Brompton ruling. Notably, the Court did not embrace this criterion in design law in its Doceram decision either, even though the AG did mention it (para. 67 AG opinion in Doceram).
Another area of penumbra is the absence of a list of factors, although some are mentioned. The Court did not do this in Doceram either, probably because it does not want to be bound by an exhaustive list and prefers to leave this assessment to the national courts. This may be wise, but ultimately could result in some disharmony and further references to the Court.
A final small regret is the lack of repetition of the dangers of cumulation as explained well by AG Szpunar in his opinion in Cofemel. That said, as the Court endorsed his opinion in that case, it is not too problematic that it did not do so again in Brompton. Its repetition of the main teachings of the judgment at paras. 31-34 (including the importance of the idea/expression dichotomy and merger doctrine) can also play that role.
It will be interesting to see what the Belgian court decides on remand. It is not impossible for it to find that the shape is not entirely dictated by function and this will depend on all the facts, which we are not privy to. For example, while there are trade mark cases where the shape was held to be entirely dictated by function (Lego brick, Philishave, Rubik cube), there are cases where it was not (Lego manikin). In two recent French cases, the patent and design were not on exactly the same aspects of the product so the two intellectual property rights could co-exist.
 Namely Cofemel, paras 29, 31 and 32 in which the Court refers to Infopaq I, paras. 37 and 39; Football Dataco, para. 39 and Levola, para. 33, 35-37 and 40.
 Case C-683/17, Cofemel, para. 50 and Opinion of AG Szpunar, para. 55; C-395/16, Doceram, para. 28.
 For a longer comment, see I. Fhima “Functionality, cumulation and lessons from trade mark law: the Advocate General’s Opinion in Brompton Bicycle”, (2020) JIPLP 301.
 Cass. com., 22 Mars 2005 N° 03-16.532-, 03-18.818 (L’Oréal v Bourjois) and CA Paris, 24 January 2014 (Nectoux v les Gourmandises de la Broceliande).