On 3 June 2021, the CJEU handed down its judgment in CV-Online Latvia v Melons (with Ilešič as a reporting judge), a case involving Melons’ infringement of CV-Online Latvia’s database of job advertisements arguably protected by the sui generis right. The facts of the case are expertly described by Tatiana Synodinou in her comment on the AG’s Opinion and we refer the readers to them. In a nutshell, a specialist search engine engaging in re-use of substantial parts of the database of a job adverts website was accused of violating sui generis database right. Unlike in Innoweb (with von Danwitz as a reporting judge), this specialist search engine does not use the search function of said website but develops its own way to explore the dataset. Moreover, the user is only offered deep links.
The referring Latvian court posed two questions. First, whether the display of hyperlinks constitutes re-utilization, and second, whether the re-use of meta tags can qualify as extraction. The CJEU immediately rephrased these questions, noting that the issue is much broader because hyperlinks and metatags are “merely external manifestations, of secondary importance, of that extraction and that re-utilisation” (para. 37). As a result, the Court broadens the question to the entire infringement test.
The CJEU says that Melons does give “users access, on its own website, to job advertisements contained in [CV-Online Latvia’s] database […].” (para. 34). So, it does reutilise the database’s content and by the fact that it previously copies and indexes the content of CV-Online Latvia’s database, it also extracts it. However, the CJEU distinguishes this case from Innoweb’s (para. 33). In short, in the Court’s view, Melons appropriates less than Innoweb’s Gaspedaal. Following Advocate General (AG) Szpunar, this conclusion is not necessarily based on how much Melons takes from the database, but on how it re-uses it (para. 33, AG Opinion).
Therefore, the CJEU concludes that these acts fall under art. 7(2)(a) and (b) of the Database Directive (the directive) but constitute infringement only “provided that they have the effect of depriving that person of income intended to enable him or her to redeem the cost of that investment” (para. 37). This important caveat introduces the raison d’être of the database protection into the infringement test. For the Court, its basis is recital 42 of the directive stating that the infringing acts must cause “significant detriment” to the database maker’s investment.
However, the Court does not stop there. Following AG Szpunar, it notes that (para. 41 of the judgment, and paras 3 and 43 of the AG Opinion):
“it is necessary to strike a fair balance between, on the one hand, the legitimate interest of the makers of databases in being able to redeem their substantial investment and, on the other hand, that of users and competitors of those makers in having access to the information contained in those databases and the possibility of creating innovative products based on that information”
Citing the AG, the Court says that “the main criterion for balancing the legitimate interests at stake must be the potential risk to the substantial investment of the maker of the database concerned, namely the risk that that investment may not be redeemed” (para. 44).
Returning to the case at hand, the Court notes that “aggregators contribute to the creation and distribution of products and services with added value in the information sector. By offering their users a unified interface enabling them to search several databases according to criteria relevant to their content, they allow the information on the internet to be better structured and to be searched more efficiently. They also contribute to the smooth functioning of competition and to the transparency of offers and prices” (para. 42).
Finally, the Court notes that competition law remains untouched and that it is up to the Latvian court to apply these criteria to the case at hand.
The decision represents a significant shift from the previous case-law. The Court requires that all acts of extraction and re-utilization must lead to a risk that the database maker is not able to recoup its initial investment because of these actions. Moreover, while considering the risk, the national courts must balance the interests of other parties as part of the infringement test. The court explicitly mentions the legitimate interests of “users” to have access to information contained in the database and “competitors” to create innovative products based on that information. The CJEU points out that if a website creates a new, better, product which enhances competition (e.g., price transparency for consumers), there is no infringement unless the risk to the database maker’s initial investment outweighs these considerations.
The judgment has a different flavour from many of the previous database judgments the CJEU has handed down, especially on the breadth of the extraction and reutilisation rights. The accent is more on a balance to be found between on the one hand the rights of the database maker to recoup its investment and on the other hand legitimate interests of competitors and other third parties, e.g., users. The interpretation of the term ‘obtaining’ already had this effect in the first 2004 decisions on the sui generis right but since then the CJEU’s tone on the rights was, as in copyright law, for an expansive interpretation (see Husovec’s 2014 criticism of the Innoweb ruling). This judgment thus breaks this trend and gives a more measured and mature view of the right. The decision fully follows AG Szpunar’s opinion and is a further confirmation of his strong influence on the copyright case law of the Court (see Derclaye’s 2020 empirical study).
It will have to been seen how the balancing requirement is operationalized by the national courts. The two immediate areas of impact are: a) scope of rights and b) evidentiary burden.
Firstly, the court clearly shows how competition concerns can be reconciled within the protection. However, this does not mean that Magill-like cases will always resolve as non-infringement. The court’s analytical framework rather seems to curb the ability of database makers to rely on investment protection in low-risk (to database makers) and high-gain (to everyone else) scenarios. Therefore, very innovative products producing strong consumer benefits or socially important re-use of public data by journalists that does not have significant impact on the investments made are likely to prevail. However, it is unlikely that the courts would operationalize the test as denying infringement in high-risk and high-gain cases. Any potential abuse in these scenarios would likely remain in the provenance of “big” competition law. Finally, high-risk and low-gain cases of parasitic or close to parasitic products, as the court seems to interpret its Innoweb ruling, will not experience any change either.
Secondly, another important effect of the CJEU’s judgment is to strengthen further the link between substantial investment and substantial part in the infringement test. This is because even if the burden has always been on the right holder to prove that a substantial investment has been taken by the defendant, it will now need to prove that there has been a significant detriment to its investment or a risk of such detriment. This might be harder to prove. The ‘significant detriment’ language of recital 42 of the database directive, which has only been cited in one of the 11 cases on the sui generis right, namely in British Horseracing Board v William Hill, and to a lesser effect, is now the centre of the inquiry into the infringement test (para. 39). In addition, this detriment to the investment is now only the main (para. 44) not the sole criterion to determine if there is infringement of the sui generis right. As a consequence, the national courts will have to engage with the evidence of investment much more than they did before.
To conclude, the decision is a welcome development. Besides AG Szpunar’s expert influence, the case confirms Ilešič’s strong preference for counterbalancing of harmonization expansion (leading often to rights expansion) with flexible concepts, such as trademark functions or risk to investment, to avoid rigid outcomes. It shows the court’s willingness to tackle other interests, including competition concerns internally, i.e. within the remit of the sui generis database right, thus avoiding over-reliance on the external tool of competition law. Para. 41 of the judgment also opens doors to the assessment of fundamental rights, such as freedom of expression and of information, on the level of infringement test. The main challenge is now on the shoulders of national judges: how will they operationalize the balancing requirement in practice? As in other domains, where the CJEU gives relatively vague criteria or non-exhaustive lists (e.g. Brompton, Doceram), we may see national courts refer yet more questions to the CJEU or we may see a body of national case law developing harmoniously. One way or another, it will be interesting.
 We only link to one as a representative of the other two cases delivered the same day by the CJEU.
 See above British Horseracing Board and Fixtures Marketing decisions.