Before the Gatekeeper Sits the Law. The Digital Markets Act's Regulation of Information Control

Printer-friendly version

Keywords: Digital Markets Act – gatekeepers – information – data – contestability – fairness.

The entry into force of the Digital Markets Act (DMA) introduces a novelty into the legal lexicon.[1] A new term elbows its way into the bustling debate over the regulation of the digital sector: gatekeeper.[2] But behind the apparent originality of the word in this context, the targets of the DMA seem to remain the same: the digital players who over the last years consistently monopolized, if not a market, at least the attention of European regulators. The plurality of terms with whom such players are designated reflects their multifaceted nature. The name Tech Giants references the dimensions of GAFAM, judged alarming by many;[3] the appellation Big Tech, with its assonance with “Big Oil” or “Big Tobacco”, suggests dominance; the concept of (multi-sided) platforms unveils their roles of intermediaries. And while some of these options lack the rigour that is expected from legal terminology, EU law is not devoid of alternatives. Against this background, what is the rationale for the introduction of the concept of gatekeeper in the DMA? What differentiates it from pre-existing nomenclatures hanging on GAFAM? Although the text of the DMA offers little more than ambiguous justifications and opaque hints, certain elements invite an investigation of the pivotal role played by information control in determining the contestability and fairness of the digital sector.

In the DMA, gatekeepers represent a specific subgroup within the wider group of undertakings providing a core platform service.[4] Undertakings are defined in accordance with the famous opinion of Advocate General Jacobs in Albany, consolidated in a Commission Notice of 1998 and embodying a fundamental pillar of the competition law lexicon.[5] The notion of core platform service, on the contrary, is another novelty of the DMA. No explicit definition is provided, but art. 2 contains a list of ten services that fall into the category.[6] Despite the heterogeneity of the items listed, core platform services share some common characteristics considered worrying by EU law-makers. Extreme economies of scale, very strong network effects, multi-sidedness accompanied by a significant degree of dependence on the service by both business and end users, lock-in effects, lack of multi-homing by end users, and data-driven advantages, are all contributing factors to the emergence of gatekeepers among undertakings offering core platform services.[7] The latter find themselves in the position of being “an important gateway for business users to reach end users”.[8]

Such a particular stance is well described by the literature on multi-sided platforms; the EU regulator already referred to the equivalent characteristics held by “digital platforms” and “digital intermediaries”.  The recourse to the term “gatekeeper”, nonetheless, hints at the reality of guarded access. Undertakings providing core platform services are in a position to control business users’ access to end users. But most of the obligations imposed on gatekeepers do not merely aim at opening up the gates connecting the two sides of digital platforms. While some provisions will indeed improve data flows between business and end users, the underlying assumption pervading the DMA appears wider in scope: contestable and fair markets in the digital sector require restrictions on gatekeepers’ control over information.

Limits on data sharing and interoperability hinder contestability. Access to data constitutes a powerful barrier to entry in the digital sector: gatekeepers can leverage their large datasets from one area of activity to another, practically foreclosing entry to new market players. In most scenarios, innovative and more efficient newcomers could hardly compete against an incumbent that has access to a large amount of data.[9] Gatekeepers indirectly control digital players’ access to information by setting technical standards and protocols for the use of their services. Restrictions to data portability and data exchange through effective interoperability have the effect of raising end users’ costs of switching. Consequently, fewer players can compete in the market. Business users entering markets characterized by a low degree of interoperability, moreover, incur greater entry costs. Their initial investment could hardly be recouped by switching to a different core service provider: they are de facto locked in the gatekeeper ecosystem. Less contestable markets give gatekeepers more room to impose unfair conditions. Contestability and fairness are intertwined.[10] In fact, the reverse is also true: unfair practices hinder the contestability of gatekeepers’ positions. An example in this sense is the intransparent and opaque information over the conditions of the use of gatekeepers’ services. One way to inhibit access to data is typically adopted by core service providers offering advertising services and consists of the disclosure of information exclusively in an aggregate form.[11] Business users do not dispose of the necessary information to compare gatekeepers’ performances against providers of similar services. This unfair practice inhibits the contestability of the market.

The implicit link between information control and the goals of the DMA (contestability and fairness) explains the choice of the term gatekeeper. Communication studies have developed an interesting literature around the so-called “gatekeeping theory”; some elements of the theory, pertaining to the context of information, are certainly applicable to the digital sector and can constitute a useful tool for understanding and interpreting the DMA. In particular, gatekeeping has been broadly defined as a process of controlling information, encompassing activities of selection, addition, withholding, display, channelling, shaping, manipulation, repetition, timing, localization, integration, disregard, and deletion of information.[12]

The overwhelming majority of obligations listed in arts 5 and 6 of the DMA relate to the various activities through which gatekeepers control information.[13] Art. 5(2) prevents addition by limiting gatekeepers’ ability to combine data from their different area of operations; arts 5(3), 5(4), and 5(5) prohibit channelling by granting that communication and trade among business and end users may take place outside of a gatekeeper’s platform, or within its platform but through software developed by business users; art. 5(6) limits withholding of information, making sure that gatekeepers do not prevent users from raising issues of non-compliance with any relevant public authority; art. 5(7) prevents selection by forbidding gatekeepers’ to require business users to adopt or interoperate with specific identification, web browser or payment services; art. 5(8) prohibits repetition by impeding gatekeepers’ ability to require end users’ registration to more than one core platform service; arts 5(9) and 5(10) prevent withholding, display, and manipulation of information related to the price and performance of advertising services.

Similarly, art. 6(2) prohibits gatekeepers from using non-publicly-available data generated by business users when competing with the same business users, thus addressing integration of information; art. 6(5) makes it unlawful for gatekeepers to self-preference their products/services in rankings, or to display information in a deceiving manner; art. 6(10) prohibits gatekeepers’ withholding of end users’ data, and mandates sharing with business users of the data generated through their interaction with end users on the gatekeeper’s platform.

Obligations contained in arts 5 and 6 differ to the extent that the latter are “capable of being further specified”.[14] In practice, this means that the Commission may supplement the obligations listed in art. 6 based on a market investigation.[15] Market investigations by the Commission can also be opened (on its initiative) to pursue two additional objectives: identifying new, or rising, gatekeepers (art. 17), and identifying new services that may be added to the list of core platform services (art. 18). These provisions are meant to preserve the relevance of the DMA in a fast-paced and constantly evolving environment. The aim is coherent with the dynamic nature of information gatekeeping.[16]

The evolution of the system (or ecosystem) comprising current and future gatekeepers and the “gated” businesses and end users is treated by the DMA as an exogenous variable. The acknowledged dynamism is unpredictable in direction, and the Commission is given the powerful role of watchdog of future innovation. Nonetheless, Network Gatekeeping Theory suggests that information gatekeeping mechanisms evolve based on the dynamic relationship between gatekeepers and the “gated”. A complete understanding of information gatekeeping requires focusing on the “gated” as much as on gatekeepers. Four variables, in particular, determine the salience of the “gated”: first, their political power with respect to the gatekeeper; second, their ability to produce information; third, the frequency and intensity of their relationship with the gatekeeper; lastly, the alternatives available to them. As the salience of the gated increases, their attitude towards gatekeepers progressively passes from passive to challenging.[17] In the latter case, the “gated” might either compete directly with the gatekeeper or resort to a new service provider, making the latter competitive against the original gatekeeper.

Some provisions in the DMA are designed to increase the saliency of gated businesses and end users. For instance, several obligations imposed on gatekeepers aim at increasing the number of alternatives available to the “gated”. Arts 6(3) and 6(4) mandate the possibility for gatekeepers’ end users to uninstall any pre-installed software or operating system, change default settings and install any third-party software applications or applications stores. Similarly, art. 6(6) protects end users’ ability to switch between services and applications other than the ones supplied by the gatekeeper. Nonetheless, little attention is devoted by the regulator to the potential differences in salience among gated users. Context-dependent variations in the reach and effectiveness of DMA’s obligations remain unaccounted for.

Lastly, gatekeeping theory suggests that perceptions matter in designating gatekeepers. As Barzilai-Nahon puts it, “being a powerful entity does not, necessarily, make one a gatekeeper”.[18] Gatekeepers designated under the DMA, however, necessarily enjoy considerable economic power.[19] They can leverage such power to establish unfair practices and conditions, undermining the contestability of the digital sector. The parallelism with the structural condition of dominance and its abuse, or at least with the typical anti-competitive behaviours subject to EU competition law, appears striking: however, the DMA is declared to be complementary to arts 101 and 102 TFEU in that it contains a list of obligations applicable ex ante, irrespective of undertakings’ conduct and market position.[20]

Multiple scholars have already highlighted the hybrid nature of the DMA: on the one hand, it shares common traits with EU competition law, to the point that it has been defined as a measure of competition policy; on the other, the choice of art. 114 TFEU as a legal basis, together with the presence within the text of elements of contract law, consumer protection, and data protection policy, might be read as an attempt to regulate digital markets as a public utility, even giving rise to a new field of law.[21] Entering such a hornet’s nest is certainly beyond the scope of this writing. More humbly, a characterization of the DMA as a regulation of information control contributes to the debate by offering a shift of perspective. Whether the DMA represents a tentative update of competition policy to make it fit for the information economy, or if it marks a first step towards a new branch of EU law remains an open question.

In one of the most famous novels written by Franz Kafka, an enigmatic gatekeeper sits before the law.[22] A man trying to pass through the gate guarded by the gatekeeper will spend his whole life hesitating before trying his luck. He will find out too late, his vision blurred by age and the last forces abandoning him, that the gate has been created expressly for him, and his death will make it useless. The DMA mirrors the novel with reversed roles. A new, enigmatic, term guards the “gate” and decides the applicability of the law, that is, the applicability of the DMA itself, to the (current and future) economic players.[23] The gatekeeper concept is introduced quasi ad hoc, and it is not devoid of mysterious elements.[24] The future interpretation and application of the DMA bear the burden of ensuring that this impressive regulatory effort does not lead to a Kafkian finale.

European Papers, Vol. 8, 2023, No 2, European Forum, Highlight of 26 July 2023, pp. 405-410
ISSN 2499-8249 - doi: 10.15166/2499-8249/662

* PhD Researcher in Law, European University Institute,

[1] Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on contestable and fair markets in the digital sector and amending Directives (EU) 2019/1937 and (EU) 2020/1828 (hereinafter “DMA”).

[2] The term “gatekeeper” does not refer to any pre-existing concept of EU law in general, nor competition law in particular. It is, however, present in the US based legal debate on corporate governance, where it is adopted to indicate “private parties who are able to disrupt misconduct by withholding their cooperation from wrongdoers” (R Kraakman, ‘Gatekeepers: The Anatomy of a Third-Party Enforcement Strategy’ (1986) The Journal of Law, Economics, and Organization 53). In this context, “gatekeeper liability” refers to the liability imposed on parties who, although not authors of misconduct themselves, are in a position to prevent it. Gatekeepers in corporate governance may prevent wrongdoing; in the DMA gatekeepers have a very different role, as they are identified as possible hindrances for (positive) increases in competition.

[3] The acronym GAFAM commonly refers to the major digital platforms: Google, Apple, Meta (ex-Facebook), Amazon, Microsoft.

[4] DMA cit. art. 2.

[5] Case C-67/96 Albany International BV v Stichting Bedrijfspensioenfonds Textielindustrie ECLI:EU:C:1999:28, opinion of AG Jacobs; Commission Notice on the concept of undertakings concerned under Council Regulation (EEC) No 4064/89 on the control of concentrations between undertakings 1998.

[6] The list encompasses online intermediation services, online search engines, online social-networking services, video-sharing platform services, number-independent interpersonal communications services, operating systems; web browsers, virtual assistants, cloud computing services, online advertising services, and advertising services when provided together with one of the above-mentioned services.

[7] DMA cit. recital 2.

[8] DMA cit. art. 3.

[9] Note that in regulating inter-platform competition, the intent of the DMA appears to be oriented towards stimulating the entry of new players more than fostering competition among established gatekeepers. Access to data is, in this context, one of the main obstacles to a level playing field.

[10] DMA cit. recital 34.

[11] DMA cit. recitals 33, 45.

[12] K Barzilai-Nahon, ‘Toward a Theory of Network Gatekeeping: A Framework for Exploring Information Control’ (2008) Journal of the American Society for Information Science and Technology 1493.

[13] Most of the “dos and don’ts” imposed on gatekeepers by the DMA are contained in arts 5 and 6. Specific obligations are imposed in art. 7 on providers of number-independent communication services. For the sake of brevity, this writing will not delve into these obligations.

[14] DMA cit. art. 6(1).

[15] DMA cit. art. 8.

[16] N Helberger, K Kleinen-von Königslöw and R van der Noll, ‘Regulating the New Information Intermediaries as Gatekeepers of Information Diversity’ (2015) info 50; K Barzilai-Nahon, ‘Gatekeeping: A Critical Review’ (2009) Annual Review of Information Science and Technology 1.

[17] K Barzilai-Nahon, ‘Toward a Theory of Network Gatekeeping: A Framework for Exploring Information Control’ cit.

[18] Ibid. 1506.

[19] An undertaking is designated as a gatekeeper if the following requirements are satisfied: a) it has a significant impact on the internal market; b) it provides a core platform service which is an important gateway for business users to reach end users; and c) it enjoys an entrenched and durable position, in its operations, or it is foreseeable that it will enjoy such a  position in the near future. See DMA, art. 2.

[20] DMA, recital 11.

[21] For an introduction, see H Schweitzer, ‘The Art to Make Gatekeeper Positions Contestable and the Challenge to Know What Is Fair: A Discussion of the Digital Markets Act Proposal’ (2021) Zeitschrift für Europäisches Privatrecht 503; N Moreno Belloso and N Petit, ‘The EU Digital Markets Act (DMA): A Competition Hand in a Regulatory Glove’ ELR (forthcoming), pre-print available at

[22] The reference is to Vor dem Gesetz in F Kafka, Der Prozess (1925).

[23] The parallel between the DMA and Kafka’s novel ends, of course, if one considers that the man in the novel is moved by the desire to access the law, while firms that control platforms would most likely rather escape it.

[24] On the definition of DMA as an ad hoc regulation, see N Petit, ‘The Proposed Digital Markets Act (DMA): A Legal and Policy Review’ (2021) Journal of European Competition Law & Practice 529.