20 years after the adoption of Regulation 1/2003, one of its most indispensable provisions continues to create uncertainty. Art. 3 has the goal to ensure consistency between the application of national and EU competition law. However, two decades later, this promise remains unfulfilled. Rather than providing clarity, it has become an instrument for fragmentation and legal ambiguity. This tension has surfaced most prominently in the Meta (formerly Facebook) case, revealing the structural limits of the current system and the cost of failing to clarify a clause born as a political compromise rather than a coherent legal rule.

Art. 3(1) established that where conduct has an effect on trade between Member States, NCAs and Courts must apply EU competition law, meaning Articles 101 and 102 TFEU. However, Art. 3(2) introduces an exception: Member States are not precluded from applying “stricter” national laws concerning unilateral conduct. Lastly, Art. 3(3) goes on by allowing Member States to enforce national rules with a predominantly different objective, provided they do not constitute competition law in a strict sense.

These exceptions appear reasonable at the time, however in practice, they have produced what some call an “inoperative compromise.” Primacy is indeed affirmed in Art. 3 (1), but Arts. 3 (2) and (3) muddy the waters by carving out broad and not so well defined exceptions. The distinction between competition law and adjacent fields like consumer protection or data governance becomes blurred, and this ambiguity was presented in the German Facebook case.

The German Federal Cartel Office found that Meta’s terms of service (allowing the combination of data collected) violated German competition law (Sections 19 and 32 GWB). In its decision, the FCO relied on the finding that Meta’s terms of service did not meet the requirements of the GDPR. The key issue, however unanswered, was whether the FCO was also obliged to apply Article 102 TFEU alongside national law, as required by Article 3(1) of Regulation 1/2003. According to the OLG Düsseldorf, the FCO’s failure to apply Article 102 was a procedural error but “irrelevant”. I mean, how relevant can it be the uniform application of EU law, right?

This went unchallenged by both the Advocate General and the CJEU. The Court simply noted that a GDPR infringement could serve as an “important indication” of an abuse under competition law, but avoided clarifying whether EU law should have been applied. The implications of applying EU competition law or not go beyond substance since they also trigger notification and coordination requirements with the European Commission and the ECN. Had the FCO invoked Article 102 TFEU, the Commission and other NCAs could have contributed to the enforcement process and influenced the decision. The failure to do so illustrates how bypassing EU law weakens the cooperative nature of the EU competition system.

Moreover, the another important question left open is what it means for a national provision to be “stricter.” The CJEU’s silence has left national authorities guessing. As noted in doctrinal analyses, “stricter” may refer to a more demanding standard for defining abuse, a lower threshold for establishing dominance, or even broader concepts of economic power like dependency or gatekeeper roles. The FCO and the OLG Düsseldorf took the view that Section 19 GWB is a competition rule (not another rule different from competition law), and that it is “stricter” because it reflects a unique German concept of protection not yet mirrored in EU case law. However, if national provisions with different substantive content are accepted as “stricter,” Article 3(2) becomes a gateway for regulatory fragmentation. There is a real risk that identical cases could be brought under national and EU competition law, leading to inconsistent outcomes.

This issue is far from isolated. In France, for instance, ARCOM recently proposed a national law to protect minors from harmful online content. Although its objective is legitimate, it touches upon an area fully regulated DSA, a directly applicable EU regulation. While this is not a competition issue, it reflects a similar dynamic: Member States asserting national standards in areas already harmonised at EU level. This growing tendency to renationalise calls urgently for judicial guidance.

The Commission could initiate, in theory, infringement proceedings under Art. 258 TFEU against a Member State whose NCA fails to comply with Art.3(1) of Regulation 1/2003. However, while this legal instrument exists, it is barely used in practice for isolated infringements by independent authorities. Moreover, such a procedure would not directly affect the outcome of the national authority’s decision, which would remain valid under national law. Clarification by the CJEU is therefore essential. The application of Art. 3 should not remain a discretionary or political matter. It involves fundamental questions of legal hierarchy, procedural coordination, and the uniform application of EU law. The Court must define the threshold for “stricter” national competition rules, distinguish them from rules pursuing predominantly different objectives, and assert when Article 102 TFEU must be applied in parallel.

The CJEU missed a key opportunity in the Facebook case to clarify the nature and scope of Art. 3. Looking forward, the next revision of Regulation 1/2003 should explicitly address this ambiguity and better delineate the boundaries between national and EU competition law. Only then can the balance between decentralisation and legal unity be truly preserved.

The European Commission’s dawn raids have long been one of the most powerful investigative tools in competition enforcement. But as companies move to the cloud, employees work from home, and communication happens across encrypted channels, we have to ask whether the dawn raid, as we know it, is still the right fit for the digital age.

In September 2024, the Commission published a Staff Working Document reflecting on the effectiveness of its investigative powers under Regulation 1/2003. Although no legislative proposals were made, the report raised questions that we can no longer afford to ignore: what happens to dawn raids when the world goes remote?

Remote Inspections: A Logical Step in a Digital World

Let’s start with a basic fact: most corporate data is now hosted in the cloud. It’s no longer locked away in filing cabinets or even in office desktops. That’s not just a matter of convenience, it’s a structural shift. And it means that the physical location of data is becoming increasingly irrelevant.

This also affects how we think about inspecting private homes. Traditionally, this has been a sensitive issue; dawn raids at domestic premises raise privacy concerns and legal complexities. But if the employee’s device is just a terminal to access a shared company server, then why go to the home at all? Access to the server is access to the data.

From my perspective, this is where remote inspections make a lot of sense. The technology is already there. Lawyers, corporate executives, and Commission officials all now use secure digital platforms in their daily work. The legal framework should be flexible enough to adapt to exceptional circumstances, like pandemics or crises, but also to the routine digitalisation of corporate environments. Remote inspections could ensure that investigations continue—without compromising privacy or relying on outdated assumptions about where evidence “lives”.

Digital Evidence Can Disappear in Seconds: Enter Freezing Orders

One of the main challenges with digital evidence is its fragility. Unlike paper documents, electronic records can be deleted or altered in seconds, even unintentionally. A recent case makes this clear: in 2024, the Commission fined International Flavors & Fragrances €15.9 million after an employee deleted WhatsApp messages during a dawn raid while investigators were already on-site.

This is where freezing orders come into play. Already discussed in the context of the DMA, these orders would allow the Commission to remotely lock down specific data sources at the start of an investigation. It’s a way to preserve evidence without stepping into physical spaces or risking its destruction.

Combined with remote inspections, freezing orders could replace the need for intrusive, resource-heavy, and potentially privacy-invasive actions, while still securing access to the full scope of relevant documents.

A Balanced, Modern Enforcement Model

I’m not suggesting that dawn raids should disappear entirely. There will always be cases where physical presence is necessary, for example, where covert practices require on-the-ground discovery. But it’s time to acknowledge that digital tools are now strong enough to support remote inspections as a standard option in the Commission’s toolbox.

What’s more, remote inspections and freezing orders would likely reduce unnecessary burdens on companies. Instead of unannounced visits and staff disruption, companies could cooperate through secure, well-defined digital channels, knowing their rights and obligations from the start.

Conclusion: Time to Embrace the Cloud

Competition law enforcement doesn’t operate in a vacuum. It must adapt to the way companies operate and the way evidence exists in a digital world. Remote inspections are not just about convenience. They’re about respecting privacy, ensuring continuity in crisis, and using the right tools for the digital economy.

When an undertaking is the target of an investigation into anti-competitive practices, the right of access to the file and the protection of confidential information to ensure a fair procedure is a balance that can be difficult to maintain. Confidentiality rings are one of the tools in competition law that try to reconcile these different interests. These procedural tools allow parties to access documents essential for mounting their defence, while ensuring that the confidentiality of sensitive business data is preserved. However, their effectiveness can be affected by the many challenges encountered in their implementation.

Described as a “negotiated disclosure procedure” in paragraph 96 of the Antitrust Best Practices, the confidentiality rings consist of a form of restricted disclosure of the file. Under this procedure, a party negotiates with an information provider from the Commission in order to receive the information contained in the file on its behalf through a small circle of individuals, whose composition is decided by negotiation on a case-by-case basis. They can thus be used both as screening mechanisms, allowing the identification of specific documents of particular interest to the addressee of a Statement of Objections, as well as a means of limiting access to documents which are at the same time confidential and useful for the addressee of the statement of objections. 

Although confidentiality rings in theory offer a pragmatic solution by allowing unrestricted access to the file while preserving the confidentiality of sensitive business information, their effectiveness and fairness have nevertheless been regularly challenged.

Confidentiality rings have played a crucial role in a number of cases, including Ethanol Benchmarks, in which sensitive documents provided by certain parties contained self-incriminating evidence. The Commission faced the challenge of providing access to the file while ensuring that the evidence was not disclosed to parties not involved in the settlement discussions. The Court then confirmed that written communications made during settlement discussions could not benefit from a presumption of confidentiality despite the self-incriminating statements they contained.^[1]Access to confidentiality rings is only granted subject to the conditions set out in paragraph 35 of the Settlement Notice, and the interests of the party to the settlement were sufficiently protected in this situation.

Similarly, in JPMorgan Chase and Others v Commission, the General Court reaffirmed the need for confidentiality rings in safeguarding both the right to a fair trial and the integrity of sensitive business information.^[2]Nevertheless, both the General Court and the Court of Justice on appeal rejected allegations that publication of the non-confidential version of the Commission’s decision would undermine the presumption of innocence and the reputation of the individuals involved in the case.^[3] Confidentiality is not an absolute right, and interim measures such as those requested by the parties must also meet strict conditions. 

Although confidentiality rings are crucial for ensuring the confidentiality of sensitive business information, is there not too much of a risk that the balance will tip in favour of confidentiality, to the detriment of procedural fairness?  Several shortcomings can be identified:

• Restricted access to evidence: given that only a limited group – usually economic experts, legal representatives or an external lawyer – is granted access to key documents, the ability of the parties to ensure a fair defence is at risk of being compromised. After establishing in Courage v Crehan that individuals must be able to claim damages for harm caused by anti-competitive conduct,^[4] the Court of Justice clarified in cases such as Donau Chemie and Pfleiderer, in the context of damages actions, that access to evidence is essential for the effective exercise of those rights. The Court ruled that a systematic refusal to grant access to evidence could undermine the effectiveness of European competition law and that national courts must balance the need for disclosure with the protection of confidential information.^[5] In order to guarantee the exercise of the right to full compensation, shouldn’t the relevant evidence be disclosed in full?
• Procedural delays: negotiating the extent of the information to be disclosed and the confidential treatment of the data can be time-consuming, potentially slowing down the proceedings and harming the parties’ ability to defend themselves. 
• Burdens: although confidentiality rings are essential to protect sensitive business information, they also impose strict obligations on the parties involved, putting their own liability at stake for breaches.

            In short, confidentiality rings are a vital tool for ensuring the delicate balance between the right of access to the file and the protection of sensitive business data. Nevertheless, as the case law of the Court of Justice reminds us, the right of access to the file remains fundamental to guarantee the parties involved a fair defence. If the procedure surrounding confidentiality rings is too restrictive, it risks compromising the very fairness it is intended to protect and, in so doing, becoming an obstacle to justice. Their application must accordingly be closely monitored in order to guarantee both procedural fairness and the right to a full defence.

---

^[1] Order of 10 September 2019, Lantmännen and Lantmännen Agroetanol v Commission, C–318/19 P(R), EU:C:2019:698, para. 52.

^[2] Order of 25 October 2018, JPMorgan Chase and Others v Commission, T-420/18 R, EU:T:2018:724, para. 38.

^[3] Order of 21 March 2019, JPMorgan Chase and Others v Commission, C‑1/19 P(R), EU:C:2019:230, paras. 32-35.

^[4] Judgment of 20 September 2001, Courage and Crehan, C-453/99, EU:C:2001:465, para. 26.

^[5] Judgments of 6 June 2013, Donau Chemie e.a., C‑536/11, EU:C:2013:366, paras. 30-32, and of 14 June 2011, Pfleiderer, C-360/09, EU:C:2011:389, paras. 30-31.

When an undertaking engages in anti-competitive behaviour, it may be necessary to intervene before a final decision is made, finding the infringement and imposing a sanction. Article 8 of Regulation 1/2003 addresses this problem by allowing the Commission to take interim measures to prevent serious and irreparable damage to competition. However, despite its usefulness, this tool has rarely been applied to date. Why is this, and how can the Commission make full use of it?

Article 8 of Regulation 1/2003 lays down three cumulative conditions for the Commission to be able to take interim measures on its own initiative:

1. A prima facie infringement: the Commission must prove the appearance or likelihood of an infringement, without conducting an analysis as detailed as for a final decision. 
2. Serious and irreparable harm: there must be a risk of serious harm to competition, leading to a change in market that is difficult to reverse.
3.  Temporary measures: Article 8(2) specifies that these measures are temporary and renewable where necessary and appropriate.

Prior to the adoption of Regulation 1/2003, the Commission could only take interim measures at the request of complainants in accordance with Article 3 of Regulation No. 17. However,in Camera Care, the Court of Justice recognised that the Commission should be able to take interim measures which are “which are indispensable for the effective exercise of its functions and, in particular, for ensuring the effectiveness of any decisions requiring undertakings to bring to an end infringements which it has found to exist”^[1]. In other words, by recognising this implicit power, the Court sought to preserve and strengthen the effectiveness of European competition law.^[2]

While the Commission has applied these measures on several occasions, paradoxically, the introduction of Regulation 1/2003, which formally granted it the power to take such measures autonomously, has resulted in only one use so far in the Broadcom case^[3]. Accused of abusing its dominant position by imposing exclusive purchasing obligations, Broadcom was obliged by the Commission to suspend the exclusivity clauses in its contracts. This decision also encouraged Broadcom to accept commitments under Article 9 of Regulation 1/2003.

But why are these measures, so useful in theory, so rarely used in practice? One of the reasons probably lies in the complexity of the procedure imposed by Regulation 1/2003. In order to adopt interim measures, the Commission must issue a Statement of Objections, give the undertaking concerned access to the file, allow it to submit written comments, and organise a hearing if requested. Some authors also point to other factors, such as:

• the removal of the possibility for complainants to apply for interim measures, 
• the fact that complainants can no longer request interim measures solely based on the harm caused to them, as the Commission can now only adopt such measures if they are likely to cause serious and irreparable harm to competition itself, 
• the need for the Commission to adopt a more economic approach to Articles 101 and 102 TFEU, 
• or even the ‘shock’ of the order of the President of the General Court in the IMS Health case,^[4] which suspended the Commission’s decision on interim measures.

However, there exist several ways in which interim measures could be applied in practice:

• Simplification of the procedure: consideration could be given to holding a hearing once the measures have been issued and reducing the need for written comments, particularly in cases of extreme urgency. Access to the file could also be limited to information directly related to the interim measures. 
• Lowering the threshold for intervention: some Member States have lowered the legal threshold for intervention. A simple ‘reasonable suspicion’ could suffice. Similarly, ‘serious harm’, without necessarily having to be irreparable, would give the Commission greater flexibility.
• Restoring the parties’ right to request interim measures: as provided for in Regulation No. 17, the parties could once again be given the right to request interim measures from the Commission and challenge any refusal, which could make this mechanism attractive once again.

In 2017, Commission Vestager stated that “if you have a tool, you should of course ask yourself why it is never used”^[5]. With the emergence of fast-moving digital markets, the use of interim measures conferred on the Commission by the DMA seems all the more useful and necessary to nip anti-competitive behaviour in the bud. If the Commission really wants to live up to its ambitions, its toolbox must not remain closed.

---

^[1] Order of 17 January 1980, Camera Care Ltd v Commission of the European Communities, Case 792/79 R, EU:C:1980:18, para. 18.

^[2] Despoina Mantzari, “Interim Measures in EU Competition Cases: Origins, Evolution and Implications for Digital Markets”, CLES Research Paper Series 1/2020, p. 1-22, at p. 5. 

^[3] Commission decision of 16 October 2019, Broadcom, case AT.40608.

^[4] Order of 10 December 2005, IMS Health v Commission, T-184/01, EU:T:2005:95.

^[5] Rochelle Toplensky, “EU considers tougher competition powers”, Financial Times, 2 July 2017, retrieved 28 May 2025, https://www.ft.com/content/7068be02-5f19-11e7-91a7-502f7ee26895.

The Directive 1/2019 of 11 December 2018 (hereinafter the NCA+ Directive) aims to empower the competition authorities of Member States to be more effective enforcers and to support the proper functioning of the internal market. It establishes minimal requirements that all National Competition Authorities (hereinafter NCA) need to comply with to ensure the effective application of EU competition law and the necessary tools to do so.

The idea of a more unified approach between NCA is not new. The European Competition Network (hereinafter ENC) was created as an addition to Regulation 1/2003 and was intended as such a forum for exchange. The NCA+ Directive aims to add to this framework to ensure closer cooperation and more harmonization.

Regulation 1/2003 has introduced a fundamental shift in the approach to competition law enforcement by decentralizing processes and giving NCA the role of co-enforcers. This shift required all NCA to have the right tools as well as sufficient powers to step into this role. Under Regulation 1/2003, the Member States had full discretion on the internal organization of their NCAs, the only obligation provided for was their cooperation with the Commission. With this change, the Commission agreed to “give up its monopoly on what was considered its ‘sharpest sword’“ (Corinna Potocnik-Manzouri) and transfer its competencies to the Member States.

The above-mentioned ENC as well as other guidances and notices were aimed to guide the NCA to more harmonization, however, as a soft-law instrument, it had only limited effects. The existing fragmentation and legal uncertainty could not be addressed effectively.

After various assessments and evaluations of Regulation 1/2003, the Commission’s 2014 Communication “Ten Years of Antitrust Enforcement under Regulation 1/2003“ found that while the decentralization was successful in increasing the enforcement of competition law infringements, not all NCAs were as effective as they should be. Therefore, the Commission presented a proposal for the NCA+ Directive in 2017, designed as a complementary instrument to Regulation 1/2003 aligning competencies. It does not introduce any substantive legal changes, and by being a Directive, Member States retain flexibility in the implementation of the requirements, instead of a “one-size-fits-all“ approach.

The measures imposed by the directive aim to cover any potential inconsistencies and ensure that NCAs have the necessary resources and competencies to take over a part of the Commission’s enforcement work. The main measures can be summarized as follows:

– NCA must be able to “exercise their powers impartially and in the interests of the effective and uniform application“ of competition law (art. 4).

– Member States need to ensure the appropriate financial and human resources for the NCA to function (see art. 5). There are no concrete numbers to determine what is “necessary for the effective performance“, so the Member States retain a certain power to determine their way of working.

– NCA must have the powers needed to gather evidence (see Chapter IV) and if necessary, impose sanctions (see Chapter V).

– NCA must have a coordinated leniency program (see Chapter VI), which encourages entities to cooperate with the authorities and present evidence of illegal cartels.

The Directive does not regulate the concrete composition or organization of the NCAs, they can be in the form of collegial decision-making bodies (like in the vast majority of Member States) but also have single-headed management (like in Poland, Germany, or the Czech Republic). This has been criticized by authors like Maciej Bernatt or Magdalena Knapp, as large differences persist between these systems. For instance, in Poland, only one commissioner is acting as a NCA, who is appointed by and acts under the supervision of the prime minister (see in this regard also the Sped-Pro case, in which the General Court ruled on its independence).

Other authors criticize the entire decentralized approach introduced by Regulation 1/2003 of which the NCA+ Directive is part. According to the Commission’s evaluations, the system is working well (although this might be a biased opinion…), as the number of enforced competition cases is higher and the Commission has the necessary air to breathe and concentrate on larger-scale infractions.

Thus, one could say that while decentralization has proven to be useful, the uniform application of EU competition law has taken a hit. NCAs often follow national particularities and competition laws might be thus renationalized. In the light of the principle of national procedural autonomy, the ENC+ Directive has only intervened in a limited manner in this area and could not bring any solutions. This leads to NCAs applying the uniform EU competition law following their national procedures, which may seem contradictory at first sight.

In conclusion, we can see that harmonization through soft-law instruments was not possible, and that even after the adoption of binding legal tools, uniform application and independence from governmental bodies are not still guaranteed. There is a balance that needs to be struck between delegating competencies to NCA, while ensuring that EU law will be applied in the same way, guaranteeing legal certainty and foreseeability.

Overall, the Directive was an important step to harmonize NCAs and eliminate major inconsistencies, such as fine calculations and enforcement approaches. Nevertheless, as many authors state, there are still important concerns, such as the political independence in Member States like Poland, which may compromise the impartiality of enforcement.