Use and abuse of databases; the intersection between IPR and competition law
By Kaare Risung
Databases are collections of information organized in a systematic or methodical way so that it can be easily individually accessed, managed and updated. They are at the heart of e-commerce, shopping and classified businesses such as Craigslist.org, avito.ru, lebencoin.fr, subito.it, kijiji.ca, FINN.no and blocket.se and other platform businesses. Databases are also important sources of the data on which AI algorithms derive their conclusions of dependencies and relationships.
Significant investments are typically made to build commercial databases. The owner needs to find and develop a suitable structure for the database so that the desired data may be efficiently accessed. Acquiring, verifying and structuring the data itself may also be costly, in particular in the initial stages of building the database. The data will furthermore have to be stored and the larger the database, the larger the cost. In order to support the use simultaneously by a large number of user, the setup must be robust and the whole setup must be maintained over time.
The investments made in databases need legal protection. The traditional basis for such protection is copyright. Copyright protection requires however that the database itself or at least its contents are original creations, which is not typically the case. The need for protection despite lack of originality gave rise to a unique (sui generis) protection of databases. In the EU, such database protection is found in the Directive 96/9/EC ("database directive"). In Norway, such protection is implemented in the Copyright act section 24. According to the database directive Article 7(1), the maker of a database which shows that there has been qualitatively and/or quantitatively a substantial investment in either the obtaining, verification or presentation of the contents shall have the right to prevent extraction and/or re-utilisation of the whole or of a substantial part, evaluated qualitatively and/or quantitatively, of the contents of that database.
Third party Use of databases
A database may be intended for the owner's own purposes, such as the databases owned by intelligence agencies, in which case the maker puts in place mechanisms to prevent third party access to the database entirely. Commercial databases are however typically by their very nature intended to be accessed by the general public. The owner would prefer for such access to take place by users accessing directly its website hosting the database. This gives the owner full control of display of advertising on the site or various forms of user payment.
In practice however, much of the traffic to database businesses comes in via generalist search engines such as Googles. Generalist search engines do not compete with the domain specific database businesses and are indeed desired by the specialists. The database business use meta tags to optimize their visibility for search engines to find and display hits from their database for relevant queries.
The existence of several separate domain specific databases has motivated some commercial players to attempt to exploit the fragmentation by offering end users domain specific search engines, so called meta search engines. Meta search engines adapt to the search forms specific to the available websites and generally offer reference to a site at poorer quality that would be the result of a direct hit. In 2005, the courts found a Norwegian meta search service Supersøk illegal exploitation of Finn.no's real estate service finn.no/Eiendom. The legal basis was database protection and unfair competition laws. In 2014, the CJEU came to a similar conclusion in the Innoweb case (C‑202/12). The Innoweb judgement concerned a search engine that specialized in advertisements for used cars. The search results obtained by the meta search engine contained hyperlinks to the websites on which advertisements were placed. In its judgment, the Court held that Innoweb reutilized, within the meaning of Article 7(1) of Directive 96/9, the whole or a substantial part of the third party databases by a means other than that intended by its makers. Innoweb thus came close to a parasitical competing product as referred to in recital 42 of the database directive, because it resembled a database without itself having data.
While generalist search engines are most often required and not disruptive to business and meta search engines have been found to be parasitical use in violation of the DBD, the latest arena of contention is a third category of search engines – content aggregator search engines. Such engines index sites within its area of focus and keep a copy of their content on its own servers. Searches thus do not use the search forms of the original websites and do make queries in real time. Content aggregators make their own search form, enabling users to carry out searches according to criteria of their own design. Such engines are typically used in the travel business, to find hotel rooms, flight and rental cars across a number of alternative providers.
The CJEU is due to clarify assessment of content aggregator search engines in its handling of Case C‑762/19 - SIA ‘CV-Online Latvia’ v. SIA ‘Melons’. The matter concerns services for employment services. Advocate General Maciej Szpunar provided his opinion on this matter 14 January 2021. Read it here. It is not entirely clear from the opinion whether SIA Melons service KurDarbs.lv is actually closer to a meta search engine or a content aggregator search engine, but it is no matter interesting to note that Szpunar views the matter as a question of whether the solution adopted by the Court in connection with ‘meta search engines’ can be applied by analogy to internet content aggregator (section 4). We note that Szpunar invites the CJEU to consider two aspects of search business: 1) the relevant technical issues vs economic effects and 2) intersection of sui generis IP protection and competition law. As Szpunar formulates the first question: The national courts should therefore verify not only whether the extraction or reutilisation of the whole or a substantial part of the contents of a database has taken place and whether it is shown that there has been a substantial investment in either the obtaining, verification or presentation of those contents, but also whether the extraction or reutilisation in question constitutes a risk to the possibilities of recouping that investment. Only where that was so should the makers of databases be entitled to prevent the extraction or reutilisation of the contents of their databases (section 47). This question puts a lot more weight to the investments made than what is evident from the wording of the DBD. DBD 7(1) states that databases were substantial investment have been made shall have protection from extraction and re-utilisation and says nothing about such protection being limited to recouping investments. Such limitation would be an unwarranted cap on returns, which could stifle investments in databases. Szpunar does not refer to an assessment of legitimate interest in DBD article 7(5), but invites the court to introduce a new limitation to the sui generis protection of databases. It will be exciting to see if the CJEU selects to go down this route.
Abuse of databases
It is further of considerable interest and perhaps not surprising, the emphasis Szpunar puts on the role of competition law in e-commerce. Note in this regard section 52 of the opinion: In fact, although that sui generis right takes the form of an intellectual property right, its origin lies in the law on unfair competition. Its objective is to protect makers of databases against the practice characteristic of unfair competition that is parasitism. However, it seems to me that that protection cannot lead to a different type of anticompetitive conduct, namely abuse of a dominant position. Yet the protection of database makers by the sui generis right might well lead to such practices and, moreover, the drafters of Directive 96/9 were aware of that risk. The desire to avoid that risk is therefore expressly stated as the raison d’être of the rule laid down in Article 13 that that directive is to be without prejudice to national or EU competition law. In so doing, Szpunar makes reference and draws the lines back to the landmark cases Magill (Joined Cases C-241/91 P and C-242/91), Bronner (Case C-7/97) as well as IMS Health (C‑418/01).
Although it is to be expected that the CJEU will continue to clarify the application of competition law to IP rights, it is hard to see C‑762/19 being fit to develop this case law in a principled manner. The national court would first have to find SIA enjoying a dominant position, which would in particular assume the ability of SIA to act monopolistically over time, which is often hard in e-commerce markets with very small barriers to entry. The opinion does not refer to any market shares, but several other sites are mentioned. The national court would further have to conclude that SIA abuses such dominant position. The advocate general implies that the potential abuse would be for SIA to use its dominant position to exclude Melons from a new market. Szpunar distinguishes between 1) the principal market for job ads on the internet and 2) a secondary market of aggregators of job ads. This approach seems entirely too simple. Such analysis requires a rigorous market analysis and justification of what value is added other than simply providing a more comprehensive search result, which is exactly what the CJEU has already consoling to be illegal poaching. Szpunar goes even further by suggesting that the consequence of abuse would be not only the obligation to provide license to its content, but denying CV Online the benefit of protection by the sui generis right provided for in Article 7 of Directive 96/9. Should this really be the outcome then we are up for some interesting battles.
We encourage our readers to carefully study he judgment of C‑762/19 once published, and in particular note 1) how the court defines the distinction between meta searches and content aggregators, 2) whether CJEU selects to introduce investments made in databases not only as condition of enjoying sui generis IP protection, but also as a limit to that protection and finally if 3) lack of providing access to third parties may have the consequence of denying sui generis right altogether.