Few issues in the field of competition law better illustrate the tension between national procedure autonomy and the application of European Union (“EU”) law by the European Commission (the “Commission”) than the question of Legal Professional Privilege (“LPP”). This tension is particularly evident in the context of inspections carried out by the Commission, where the applicable standard protecting lawyer–client communications may be less protective than those provided by the laws of certain Member States.

Neither Regulation 1/2003 nor the previous (Regulation No 17), contains a specific provision governing lawyer – client communications. Nevertheless, the Court of Justice in AM & S Europe[1] later confirmed in Akzo Nobel[2], held that Regulation No 17 must be interpreted as providing protection only for correspondences with independent lawyers—that is to say, a lawyer not linked to the undertaking by an employment relationship and who are members of the bar in a Member State, where such communications occur for the purposes —where such communications relate directly to the interests of the client’s rights of defence.

As a result, correspondences with in-house lawyers, even if they are fully qualified and acting in a professional legal capacity, are excluded from this protection under EU law. The Court’s rationale is based on the premise that-in house lawyers, due to their employment relationship, cannot be presumed to act with the same degree of independence as external counsels and hence they cannot ignore the company’s commercial strategy. Accordingly, communications between a company and its own legal department are not shielded from seizure or review by the Commission during an inspection, even when these involve confidential legal advice

This position stands in contrasts to the broader protection recognised in some Member States — such as those of Ireland, Poland, Portugal and the Netherlands — where national rules on legal privilege are more expansive and extend, under certain conditions, to communications with in-house counsel[3]. This discrepancy between EU’s narrow definition and the broader national standards creates a structural asymmetry: what may be privileged in a domestic investigation might not be in an EU led one.

This disparity raises and important problem in terms of legal certainty and rights of defence. Undertakings subject to competition investigation may find their procedural rights diminished depending on the authority they are being investigated. Furthermore, more worryingly, National Competition Authorities, aware of their own legal limitations, may be tempted to allocate the case to the Commission via the European Competition Network Notice on the basis that the Commission is “better placed to act”[4]. In doing so a conduct could result in an infringement due to the evidence obtained under EU law that would otherwise be inaccessible under national law.

While it is true that this is a procedural convenience and enhances enforcement capabilities, it becomes problematic when it touches fundamental rights of defence. The ability of a national authority to circumvent the national privilege protections by strategically triggering the involvement of the Commission risks undermining the rights of defence of the parties. Hence, if such situations were to occur, individuals and undertakings would be unable to predict the how to reliably assert their rights. Thus, in practice, whether a communication is protected or not would no longer depend on its legal character or purpose, but rather on the strategic choices of the authorities.

To remedy this, a uniform or harmonised standard on legal professional privilege across the European Union should be considered.

---

[1] Judgment of 18 May 1982, AM & S Europe Limited v Commission of the European Communities, C- 155/79, ECLI:EU:C:1982:157.

[2] Judgment of 14 September 2010, Akzo Nobel Chemicals Ltd and Akcros Chemicals Ltd v European Commission, C-550/07 P. ECLI:EU:C:2010:512.

[3] Latham & Watkins LLP, Pro Bono in the European Union (2012), https://www.lw.com/admin/upload/SiteAttachments/pro-bono-in-the-european-union.pdf.

[4] European Commission, Commission Notice on Cooperation within the Network of Competition Authorities, OJ C 101, 2004, para. 9.

Where you would want to apply interim measures is where you have novel cases and the problem of doing interim measures… is that you still have to prove that it is needed[1].

In the landscape of EU competition enforcement, interim measures have long been the exception rather than the norm. While theoretically powerful, their practical use has remained limited—so much that between IMS Health (2001) and Broadcom (2019), the European Commission has not adopted a single interim measure decision. This 18 year gap raises an important question: why has the Commission been so reluctant to use this tool?

At its core, an interim measure is a provisional remedy: it is meant to prevent serious and irreparable harm to competition before the final decision in a case is adopted. It works as a safeguard against the irreversible consequences that could result if the allegedly anticompetitive conduct were allowed to continue unchecked during the investigation.

Although in Regulation 17 the Commission was not explicitly granted such powers. The power to impose interim measures was originally derived from the case law, indeed, in Camera Care the CJEU held that the EC had the implied power to order interim measures to prevent infringement decisions from becoming ineffectual or even illusory because of the action of certain undertakings[2]. This was later codified in Article 8 of Regulation 1/2003, which formally enables the Commission to impose interim measures ex officio—that is, on its own initiative. This also meant that undertakings no longer enjoy a right to request for such measures directly.

Several factors help explain the Commission’s longstanding reluctance:

• First, the removal of the complainant’s right to request interim measures has reduced the pressure on the EC to act. As suggested in paragraph 80 of the Commission’s Complaints Notice, complainants should seek interim relief to national courts or NCAs[3]. In practice, this has shifted the burden of interim enforcement to the national level.
• Second, interim measures require a parallel procedure with distinct procedural guarantees and the likelihood of judicial review. For a resource-limited authority like DG COMP, this implies added complexity and risks delaying the main investigation.
• Third, the Commission must show not only a prima facie infringement, but also the existence of a serious and irreparable harm to competition. This requires a high evidentiary standard, particularly difficult to meet in cases involving new theories of harm.
• Fourth, if interim measures are later found unjustified, the Commission may face damages (Schneider Electric). This can act as a deterrent effect and it can be magnified in cases involving new legal theories of harms, where the danger of a false positive is greater.
• Fifth and lastly, there’s the concern of false positives, where early intervention, especially in complex or innovative markets, risks “picking winners.”

Yet in Broadcom (2019), the Commission’s approach shifted. The Commission imposed interim measures to order Broadcom to stop imposing exclusivity requirements to purchasers—its first use in nearly two decades. Interestingly, these interim measures later formed the basis for a commitments decision under Article 9. This suggest that interim measures can function as a form of strategic leverage to prompt voluntary compliance under Article 9.

This evolution could indeed point to a more pragmatic approach: while still rarely used, interim measures may become a tool of strategic negotiation for future commitment and therefore, to address the concerns of the enforcer in a more direct and effective way. In this sense, interim measures may be awakening, not as blunt instruments, but as subtle tools for anticipatory enforcement.

---

[1] Alec J. Burnside and Adam Kidane, “Interim Measures: An Overview of EU and National Case Law,” e-Competitions Bulletin, June 7, 2018: https://www.concurrences.com/en/review/issues/no-4-2016/case-comments/evidences-the-general-court-of-the-european-union-holds-private-wiretaps

[2] Order of the Court, 17 January 1980, Camera Care Ltd v Commission of the European Communities, para. 2, ECLI:EU:C:1980:18.

[3] Commission Notice on the handling of complaints by the Commission under Articles 81 and 82 of the EC Treaty.

With the adoption of Regulation 1/2003 and the ECN+ Directive the system of competition law enforcement was meant to become more efficient and fairer. But the question is how it plays out in practice.

Previously, in the field of leniency applications, companies were facing difficulties when submitting their leniency application resulting in the fact that whoever approaches the Commission first, would win. If in the meantime another undertaking was quicker in applying for leniency before a NCA, the undertaking could only hope that the Commission would pursue the case or that you had not given evidence that would not benefit you. This has gladly been resolved with the ECN+ Directive. Now the undertaking can file a summary application to secure its place in the line before the NCA but does not have to provide the evidence straight away.

Still, some inconsistencies remain in the decentralised system.

For example, NCAs are not allowed to issue negative decisions as the Court has held in Tele2Polska. If this would be the case, it is argued that the uniform application of EU law would be undermined. But what value does the decision of the NCA have, to conclude that there are no grounds for action under Article 5 of Regulation 1/2003? One could even question the effet utile of Article 5. So even after the NCA has found that the undertaking has done nothing wrong, the company will never be sure to be spared of any further investigations by the Commission as the decision does not cause any legitimate expectations for the undertaking (C-681/11 – Schenker and others, para. 42). 

If a NCA has deliberately decided not to further proceed with the investigation, why should this assessment be countered without any new evidence being brought forward? This interferes with the constitutional principle of legal certainty.

Furthermore, if the Commission finds that the case bears so much relevance, it is free to make use of its powers under Article 11 (6) of Regulation 1/2003 and take on the case itself.

To establish more legal certainty, the NCA should be obliged to contact the Commission before rendering its decision, just like under Article 11 (4) of Regulation 1/2003. In that case, no expectations would have been created for the undertaking causing less interference with the principle of legal certainty.

Another disadvantage for companies is the variation of the procedural safeguards in EU and national competition law. 

For example, the Commission is not obliged to wait for a lawyer when interviewing staff during an inspection. This is justified by the fact that there is no incrimination for individuals under EU competition law. But what happens if the Commission for some reason closes the case and the NCA takes up on it using the evidence the Commission has obtained during such an interview? In some Member States, such as Germany, incrimination of individuals is possible. Although Article 12 (3) of Regulation 1/2003 states that exchanged information can only be used as evidence if information has been collected in a way which respects the same level of protection of the rights of defence as under national law, the information nevertheless is out there. 

One could say, this is simply the risk of a decentralized system. But this issue could easily be remedied by the obliged assistance of a lawyer. In case a lawyer cannot actually be present, virtual assistance or assistance of a lawyer by phone would already be an improvement regarding the procedural rights. This should become an obligation for the Commission to respect as a procedural safeguard.

A further problem regarding legal certainty for businesses is displayed by the Amazon Buy Box case. The Commission investigated the conduct of Amazon but carved out the Italian market. The Italian NCA could therefore proceed with its own investigation. Having two decisions on the same behaviour but splitting the geographic market raises concerns regarding the circumvention of the ne bis in idem principle and legal certainty. This would have been prevented if the Commission had applied its very own Cooperation Notice. This again shows that the decentralised system can negatively affect enterprises.To put it in a nutshell, divergences can never be ruled out in a decentralised system. A decentralised system also has its advantages. But it should not go to the detriment of undertakings or individuals, and it should definitely not go to the detriment of the constitutional principles that lie at the heart of our legal system. Some divergences can be avoided by simply applying the existing legal framework. But some improvements for procedural safeguards are needed. The summary applications set a good example.

The principle against self-incrimination is one of the elements of the right to a fair trial under Article 6 ECHR and is also stipulated in recital 23 of Regulation 1/2003 (ECtHR, John Murray v UK, No. 1837/91). The principle entails, generally, that no person can be obliged to produce evidence against oneself, including the right to stay silent and not to answer questions. Consequently, it is questionable how this aligns with the obligation to reply to compulsory requests for information (RFI), including to hand out documents and to provide statements. 

The case law of the CJEU tried to strike a balance between the effet utile of Article 101 and 102 TFEU and the principle against self-incrimination. The Court established a different treatment between factual questions and documents that might incriminate the undertaking and answers that might entail the admission of the existence of the infringement (C-374/87– Orkem, paras. 34-35). In the case of the latter, the undertaking is entitled to remain silent. The Commission therefore tries to ask for factual information. But the distinction remains blurry as the difference between those two categories remains difficult to establish in practice. This was also implied by AG Wahl in HeidelbergCement (C‑247/14 P) when suggesting that providing factual evidence could be tantamount to admitting guilt. For example, asking if representatives of an undertaking took part in a cartel meeting could be seen as a factual question but at the same time, if the undertaking replies in the affirmative, it entails the admission of an infringement – at least indirectly. However, the Court did not see this the same way in Orkem when it held that it is unproblematic to disclose the circumstances of the meetings and the capacities of the attendees (para. 37).

As criminal law and antitrust law are not identical, the CJEU decided in Mannesmannröhren (C-411/04 P, para. 42) that the right against self-incrimination is not applicable to the same extent as it is in criminal law. However, the finding of an infringement may have severe consequences for the undertaking involved, which should be kept in mind when striking the balance. Several problems arise in this context:

First, the undertaking of course has the possibility to call the Hearing Officer and hope that he/she will side in favour of the principle against self-incrimination. But the recommendation is not binding so the definite evaluation rests in the hands of the Commission.

Second, if the undertaking still refuses to answer the question, it risks a fine. Of course, it can challenge the fine, but this should not be a counterargument in a Union based on the rule of law. The respect of the principle against self-incrimination shall not merely depend on invoking this right in judicial proceedings. Rather the Commission should ensure that the rights of the undertaking are sufficiently respected when issuing the RFI.

Third, although under EU competition law a natural person cannot be held liable, this is different in some national legal systems. So if the Commission decides to close the case for any reason after having received self-incriminating information, and then the national competition authority takes up the case, the answer of the undertaking or its representatives appears in a totally different light. Especially regarding the fact that due to the individual liability under national law, the individual could have benefitted from the right not to self-incriminate oneself.

Fourth, regarding the increased influence of right-wing parties and the backsliding of democracy and the rule of law in several Member States, clear safeguards need to be provided. Even if we trust the actions of the Commission today, we never know who might hold power tomorrow. That is why basic fundamental principles must be laid out more clearly.

This is why I agree with AG Wahl that not simply factual questions should be allowed to be posed but rather it has to be assessed whether the given answer might imply an admission of guilt. Thinking this further step – which results in a case-by-case assessment – is the least we can do in a European Union based on the rule of law as the given answers might have long-lasting and severe effects on undertakings and individuals.

On 14 May 2025, the next round of the Vivendi/Lagardère battle against Commission RFIs unfold in a court hearing in front of the GC. A perfect opportunity to recapitulate on the “what happened so far”.

In July 2023, the Commission opened an investigation of a potential early implementation (gun jumping) of the initial Vivendi/Lagardère merger, originally cleared subject to conditions (for the merger case see M.10433, the gun-jumping see M.11184). In this context, the Commission sent RFIs under Article 11(3) EUMR to the parties, which were later challenged. In the interim proceedings T‑1097/23 R (Vivendi) and T‑1119/23 R (Lagardère), the GC rejected their applications for suspension of the RFIs. Upon appeal in C‑90/24 P(R)-R and C‑89/24 P(R)-R, the ECJ VP however found that the Parties demonstrated infringements of the parties’ rights of defence by the GC’s dismissal for suspension of the RFI. Consequently, the ECJ VP ordered suspension of its execution until the GC decides further: the ECJ VP held that infringements of the rights against self-incrimination, rights to privacy and protection of journalists’ sources when assessing the urgency criterion. However, the prima facie case / fumus boni juris criterion needed still to be assessed by the GC. Then, in June 2024, the GC VP ordered the suspension of the RFIs’ execution until the end of the main proceedings, insofar as information on private life has not yet been transmitted to the Commission. Data already received by the Commission shall be kept sealed. So long, the parties are not required to further comply with the RFI. Now, a first judgment in the main proceeding may require the Commission to exercise stricter scrutiny while investigating by RFIs.

The cases have an impact on the Commission’s powers under Article 18 Regulation 1/2003 (Reg 1), in classic antitrust proceedings. While undoubtedly, the investigatory powers of EUMR and Reg 1 are based on two different legal instruments, the similarities in wording and basic requirements remain. The similarities already existed in the parallel preceding Regulations 17 (antitrust) and Regulation 4064/89 (ECMR), in their respective Articles 11. Both regulations also evolved similarly. With recasting the antitrust procedural framework into Reg 1 in 2003, and the recasting of the ECMR into the EUMR in 2004, for example, the former mandatory requirement to issue a non-binding RFI before a decisional RFI was abolished under both regulations (see Article 11(5) ECMR, or Article 11(5) Regulation 17). The standards may be applied even more by analogy to Article 18 Reg 1 because the Commission investigated finable infringements, see Article 14(2) EUMR.

Now what conclusions can be drawn and what is still open? Vivendi and Lagardère are both cases where the Commission’s search terms and their operation that had been ordered in the RFIs yielded results, which highly likely pertained to private information. In the Lagardère case, that disclosure would have entailed criminal liability under French law. In both cases, the information effectively sought by the Commission was held on dual-use devices, i.e., professional devices where private usage was allowed for, or private devices that were used at least once professionally. If certain private data (sensitive, strictly personal) was to be disclosed under compliance with the RFI, the Commission foresaw a virtual data room procedure to keep private data under limited access within the Commission. At least in the interim procedures, the Courts did not find this sufficient a protection. Suspensions had thus been ordered. The orders are remarkable: At least in interim proceedings, the Court held that search terms of “general nature” lead to encroachments upon private life, and that the virtual data room procedure cannot undo this encroachment. Whether this procedural balancing is still sufficient, remains to be seen in the judgment in the main proceedings. Additional further clarification may be made with regards to the quality of the search terms and their operation. In the Vivendi case, the ECJ considered “sensitive data”, “strictly personal data”, and a broader spectrum of data pertaining to private life. The latter, seemingly broadest criterion was retained by the Court. What precise data has been uncovered by the Commission remains to be seen, and with it potentially an overhaul of the standards how the Commission may investigate with RFIs, where businesses tend to allow private usage on corporate devices.

In the (first) final showdown, the GC’s may give important clarifications to the Commission how to operate RFI proceedings. This may then still be appealed by the Commission, going then in the final final round before the ECJ. The overall question remaining to be answered is: is a privacy exception exceptional (to these cases) or may it be generalised? To be seen soon.

While not completely overturning the Alrosa case law, the Court of Justice’s judgment in Canal+ (Case 132/19 P) crushed the “sky is the limit” reasoning that it introduced when it comes to commitments. Despite this, the weak application of the proportionality principle featured in Alrosa is still alive.

In Alrosa, the Court of Justice ruled that proportionality means something different in commitments cases (Article 9 of Regulation 1/2003), as compared to remedies imposed by the Commission in the context of infringement decisions (Article 7 of Regulation 1/2003). Accordingly, applying the principle of proportionality “is confined to verifying that the commitments in question address the concerns it expressed to the undertakings concerned and that they have not offered less onerous commitments that also address those concerns adequately” (Court of Justice, Alrosa, para. 41). When it comes to Article 7, however, proportionality has a stricter meaning, which is that the Commission’s remedies must “not exceed the limits of what is appropriate and [must be] necessary” (General Court, Case T-111/08, para. 323).

It is fair to say that for a long time, Alrosa was seen as a free pass for the Commission, as it allowed it to accept commitments that went beyond what was considered as proportionate under Article 7. In other words, the sky was the limit. However, there was always a caveat: “when carrying out [the] assessment [of commitments], the Commission must (…) take into consideration the interests of third parties” (Court of Justice, Alrosa, para. 41).

The Canal+ judgment was the first time the Court was overturning a commitments decision — and it did so on the basis on the interest of third parties. In other words, the last sentence of the 41th paragraph of the Court’s judgment in Alrosa had remained dead letter until Canal+ — and it came back to bite the Commission. In that sense, Canal+ put the Alrosa genie — which could grant the Commission almost any wish when it came to commitments cases — back in the bottle, with a clarification. In Canal+, Paramount’s commitments involved not enforcing certain territoriality clauses, resulting in Canal+’s contract with Paramount (which was based on this clauses) becoming effectively void. While the General Court claimed that disputes based on this contract could be solved before national courts, the Court of Justice dismissed this possibility, invoking Article 16 of Regulation 1/2003, which clearly states that national courts cannot issue decisions that conflict with a Commission decision. This meant the Commission adopted a decision that eliminated contractual rights of third-party Canal+ without giving it real legal recourse.

In Canal+, the Court distinguished (i) the proportionality of the commitment to solve identified competition issues from (ii) the proportionality of the commitment with regards to third parties.

While Canal+ goes a step towards a better consideration of third parties, the weak definition of proportionality for commitments themselves (i.e. with regards to their objective) remains well and alive. This poses several issues:

• Alrosa effectively shielded the Commission from substantive judicial control of the content of its commitment decisions, thereby becoming a preferred tool for DG COMP.
• Without a substantive proportionality review, the Commission can use commitments in a “regulatory” fashion, i.e. in order to influence the structure of markets (beyond what is necessary). It must be noted that although it is the parties who formally propose commitments, the Commission often has a significant influence on their content.
• Several key procedural rights apply with an Article 7 decision, and not an Article 9 decision — for example, the right to be heard and have access to the file kick in when a statement of objection is addressed to the undertaking(s), which is not the case for commitments.

While Canal+ was a step in the right direction to protect the rights of third parties, the principle of proportionality (enshrined in Article 5(4) of the TEU) has been durably weakened — even after Canal+ — by Alrosa.

In her song So What, P!nk sings that her mistakes don’t matter because she’s still “a rockstar.” Similar things might have been heard in the corridors of DG COMP after the 2017 Intel judgment, in which the Court of Justice considered the Court’s breach of the right of access to the file did not vitiate its decision (although the case was sent back to the General Court). Not all such breaches lead to annulment of a decision (or part thereof): this blog explores why, focusing on evidence from interviews.

Access to the file is a direct translation of the principle of equality of arms. In antitrust proceedings, access to the file is granted to the parties that are the addressees of a statement of objection (Article 15(1), Regulation 1/2003).

In antitrust proceedings, the basic principle is that a Commission decision (or part thereof) can be annulled by the CJEU on the basis of a procedural breach only if the applicant(s) show(s) that, absent the procedural breach, the outcome would have been different (Court of First Instance, Hercules Chemicals v. Commission, para. 56). When it comes to incriminating evidence, the parties must therefore show that the Commission based itself on the document to reach its decision and that its decision would have been different had the document not been used by it (Court of Justice, Aalborg Portland and Others v Commission, para. 73). Reversely, when it comes to exculpatory evidence, the applicant has to show that it could have defended itself differently and that this could have had an impact on the Commission’s decision (Ibid., para. 74).

The key question — for both inculpatory and exculpatory evidence — is therefore the following: what does one need to prove to show that a breach of the right of access to the file would have impacted the course of proceedings and that this could have led to a different outcome in terms of the content of the Commission’s decision?

An assessment of the relevant case law reveals a standard that takes the shape of a double-edged knife for the Commission. On the one hand, it means that many procedural breaches of the right of access to the file will be void of practical consequences, for it is sometimes fairly difficult for an undertaking to meet this standard. In the Intel case, for example, the Commission organized a meeting with Dell but did not properly disclose it or granted Intel access to the minutes — a procedural breach, according to the Court of Justice (Court of Justice, Intel, para. 96). Intel claimed the meeting contained exculpatory evidence that could have had an impact on the Commission’s decision.

However, Intel failed to prove that, based on an internal DG COMP note that was provided to it, it could show the meeting contained the said-exculpatory evidence; additionally, Intel did not make use of the possibility to ask for the person the Commission interviewed to be summoned before the General Court; and it failed to prove it contacted that person to obtain the exculpatory evidence (Ibid., para 99-102). In light of the above, the Court refused to annul the decision because of the procedural breach. Similary, in Google Android, “Google was able to obtain information from the Commission about the substance of what was discussed during those interviews (…) [but] has not put forward any detailed argument that might explain how it might have been better able to ensure its defence” (General Court, Google Android, para 946-947).

These judgments show the Court’s focus on establishing whether the undertaking can genuinely prove that the outcome of the proceedings would have been different, in order to distinguish claims that are aimed purely at discrediting a decision of the Commission, from those where the undertaking genuinely opiates that it has been harmed. The formula is the following: if the undertaking has sufficient information about the evidence but fails to prove that its defense during the proceedings (and therefore, potentially their outcome) could have been different, then the Court is rather strict and reaching the annulment of a decision is difficult.

However, when information on the content of a Commission’s meeting is so sparse that it is not possible to eliminate the possibility that exculpatory evidence might have been produced and that it could have been used by the undertaking(s) to change the outcome of a case, the standard of proof is less difficult to meet for the undertaking(s) which ‘only’ need(s) to provide arguments that “tend to demonstrate in concrete terms” that the interview “could” have contained evidence allowing for a better defense and potentially changing the outcome of the Commission’s (General Court, Quallcomm v. Commission, para. 221)

In other words, when it comes to access to file, when the Commission does it job, especially when it comes to recording and providing a proper record of the evidence produced in the course of meetings, it is very difficult for a procedural breach of the right of access to file to lead to the annulment of a decision (or part thereof). However, when a proper record has not been given access to, and that the undertaking(s) show(s), in a genuine manner, that there exists a possibility that exculpatory evidence could have been produced and that it could have helped building a better defense, with a potential favorable impact on the outcome of a case, a decision (or part thereof) can easily be annulled. Hence the double-edged sword of this standard from the Commission’s perspective.