Our European society is more connected than ever. The creation of hundreds or thousands of files per day, the storage of huge amounts of data and the democratization of the cloud impose an obligation on companies to manage this data, often containing very sensitive information related to them.

The enforcement of European competition law must nowadays take account of the challenges that have arisen in this new digital economy, especially in the use by the Commission of its investigative powers. When the institution suspects a serious infringement of competition law and a possible destruction of evidence, it can issue a decision that will allow the conduct of “dawn raids”, therefore obliging the undertakings concerned to comply and cooperate.

Article 20(2) of Regulation 1/2003 provides that during inspections, inspectors can examine books and other records related to the business, and are therefore entitled to search through laptops and other devices, having unrestricted access to them. Business records include all documents, irrespective of their format, i.e. emails can be seized as well as digital files, USB drives…¹ Moreover, Regulation 1/2003 allows the Commission “to take or obtain in any form copies of or extracts” from these records. EU officials are granted a wide margin of discretion because they can copy digital data and use innovative tools to recover deleted files. 

The challenge of determining to what extent digital data can be seized and collected during dawn raids is high. As the concentration of data through different digital platforms grows, the Commission tries to extend its access to capture all useful data for evidence gathering purposes. Assisted by its forensic team, the collection of digital data goes way further than laptops and services inside the company, but deviates to instant messages, clouds and other digital devices provided that they are used for business activities. 

The incredibly intrusive nature of dawn raids suggests that inspections covering digital data need to comply with procedural safeguards, especially those enshrined in Article 8 ECHR (respect for private life and correspondence). The Vinci Construction judgment rendered by the ECtHR in 2015² was an opportunity for the Court to reaffirm the conditions of lawfulness of  a dawn raid : it is prescribed by law, it pursues a legitimate aim and is necessary and proportionate. However, can we consider that the extension of the inspection to all data contained in company chats, private messaging conversations and private company devices is proportionate and complies with fundamental rights? It seems, in relation to this case law, that a violation would be likely considered since judicial protection against excessive investigations has been reinforced. 

Should the inspection be proportionate, any action intended by employees to delete or cover relevant data would be fined for obstruction to the investigation. This question is relevant and important because the Commission’s request to access unusual supports became part of its wide investigative powers. The nuance is that private conversations cannot be monitored by employers but, if the Commission observes an incomplete transparency in seized documents and data, the undertaking will be held liable. 

That is precisely what happened when, during an inspection at the premises of International Flavors & Fragrances in March 2023, the Commission asked to review the mobile phones of certain employees, and observed that one of them intentionally deleted WhatsApp messages (instant messages) exchanged with a competitor that contained commercially sensitive information. The company was fined an amount of €15.9 million, despite having helped to recover the deleted data. This event was a turning point in the history of European competition law as it was the first time that an undertaking was fined for having deleted instant messages on a mobile phone³. 

What to expect in the near future regarding the conduct of dawn raids involving the access and treatment of digital data? We can assume that the Commission and the national competition authorities will continue to extend the means to get access to digital data, in light of the wide powers enshrined in Regulation 1/2003. Given that leniency applications have decreased over the years and the amount of dawn raids remained stable, having access to more data while being protected from allegations of violation of fundamental rights will allow a more effective competition law enforcement. 

Nevertheless, one major issue that competition law enforcers need to consider lies in the access to the cloud. Due to the new ways of working adopted by undertakings with remote work, business data is often contained in the cloud. Employees can easily access it at home via different working applications and devices. It is however a very technical task to select the relevant information, and the Commission may require a large cooperation by inspected companies. This in turn increases the risks of obstruction to the inspection… 
That is why, in order to reduce the risks of obstruction due to the collection of extensive digital data, all employees must be aware of the scope of inspections, which includes instant messages to competitors. Furthermore, working from home during the day of the inspection does not mean that they are excluded from the obligation to give their personal device containing business information (meaning not deleting anything). The Commission, in its revision of Regulation 1/2003, has considered introducing a “freezing order” that would impose temporary restrictions on the deletion or alteration of data⁴. This would provide more clearance of inspected undertakings.

1.  L. Ortiz Blanco, European Union Competition Procedure, Chapter 8 on Inspections. ↩︎
2.  Judgment of 2 April 2015, Vinci Construction et GTM Génie civil et services c. France, ECLI:CE:ECHR:2015:0402JUD006362910. ↩︎
3.  Francesca Miotto, “European Union: dawn raids prove critical for authorities to manage dynamic markets and emerging tech”, Global Competition Review, 25 April 2025.
   https://globalcompetitionreview.com/market-review/market-review-cartels/2025/article/european-union-dawn-raids-prove-critical-authorities-manage-dynamic-markets-and-emerging-tech. ↩︎
4.  Liliane Gam, Elisha Kemp, “Hold the phone: antitrust authorities looking closely at instant messages in dawn raids”, Linklaters, 12 July 2024.
   https://www.linklaters.com/en-us/insights/blogs/linkingcompetition/2024/july/hold-the-phone-antitrust-authorities-looking-closely-at-instant-messages-in-dawn-raids. ↩︎

The Leniency Notice adopted in 2006 introduced the possibility for undertakings to apply for immunity from fines, provided that they are the “first to submit information and evidence which will enable the Commission to (a) carry out a targeted inspection in connection with the alleged cartel ; or (b) find an infringement of Article 81 EC in connection with the alleged cartel”. The prospect of obtaining full immunity is one reason for the success of the Leniency program, especially since the amounts of fines imposed for cartel infringements has increased in recent years¹. In that sense, a leniency application results from a balance between costs and benefits, meaning that it must be sufficiently incentivizing for an undertaking to want to expose the cartel. However, we observe that after assessing the risks, cartelists struggle to see a real benefit in revealing the cartel², and prefer to adopt a “wait and see” strategy, which can be reflected in the settlement procedure.

The settlement procedure is initially designed as a complementary tool to leniency, requiring undertakings to acknowledge their unlawful behaviour in exchange for a reduction of fines. Although both tools aim at cartel deterrence, they occur at different stages of the proceedings: while the application for leniency is introduced before the onset of the investigation, the settlement procedure takes place at a later stage, in order “to speed up final resolution and reduce all parties’ legal costs”³. In this sense, “settlement is a procedural efficiency instrument while leniency is an instrument to gather evidence”⁴. In that regard, the settlement procedure is useful for undertakings that colluded and decided not to reveal it, but all the difficulty in the implementation of this tool resides in the risk of undermining the effectiveness of leniency.

Consider a simple case where the undertakings, after assessing the costs and benefits to apply for leniency, decided not to reveal the cartel. The Commission then decides, post investigation, to offer them a settlement. They accept the offer. At that moment, the risk of a pro-collusive effect is high: the major challenge for the institution, through the bilateral settlement discussions, is now to set an appropriate settlement amount that does not reduce the incentive to apply for leniency. Indeed, the lower the settlement amount (i.e., the lower the fines), the higher the expected value to collude and not reveal (which leads to an increase of cartel formation and reduces the effectiveness of leniency).

The Commission therefore has to balance between the need to offer a settlement that is attractive enough for undertakings to accept, but without making it too generous as it could increase the profitability of collusion⁵. An inappropriate amount will encourage cartelists to adopt a “wait and see” strategy, i.e., not reveal the cartel in the first place⁶. This obviously contributes to another challenge of the settlement procedure, which is not to reduce the deterrence effect of the leniency program.

On that matter, the Settlement Notice provides for a reduction of fines of 10%, but this percentage cannot in any way be applied systematically to every fines reduction granted to cartelists. One criterion that contributes to the deterrence effect of the settlement procedure is the asymmetry of the settlement fine reductions between undertakings. In other words, the first undertaking to accept the settlement procedure is granted the highest fine’s reduction. This uncertainty regarding the amounts will de facto lead to a  “race to settle”, and ensure the effective enforcement of cartel cases. Moreover, it can also strengthen the leniency mechanism , because the fear of ending up with the less favourable offer will encourage undertakings to take the lead and reveal the cartel. 

However, this deterrence mechanism may be insufficient to protect against the decrease in incentives to apply for leniency, if the latter is structurally failing. We notably observe that for undertakings which would come second or third to reveal the cartel, their incentive to apply for leniency decreases, because they would only benefit from a reduction of fines. It has been revealed that in cases concluded under the settlement procedure, 85% of the parties benefited from leniency⁷. That shows that non-first leniency applicants prefer to wait and negotiate a settlement at a later stage of the procedure, rather than expose themselves at the beginning. That is why cartel deterrence needs to be supported by other active tools, especially ex-officio: in Italy, cartel enforcement remains strong particularly because of an effective cooperation between agencies, which protects incentives for companies to self-report⁸.
As a result, there is no doubt that settlement fine reductions contribute to the success of that procedure, but also impact the effectiveness of the leniency procedure. Clear responses remain to be seen and this is up to the European Commission to ensure the real complementarity of these tools which cannot lead to an exclusionary effect of one mechanism due to the wrong calibration of fines, which is clearly stated in a case-by-case analysis. This task is even more complicated since we could argue, on the contrary, that a higher amount of fine reduction of 20 or 30% would not impact the attractiveness of the leniency program, as leniency applicants benefit from immunity or reduction of fines up to 50%. Furthermore, the percentage of fine reduction of 10% set currently for the settlement procedure might not be sufficient to incentivise undertakings which have not applied for leniency, to settle⁹.

1.  Johan Ysewyn and Siobhan Kahmann, “The decline and fall of the leniency programme in Europe”, Concurrences n°1-2018, pp.44-59. ↩︎
2.  L. Ortiz Blanco, “European Union Competition Procedure”, Chapter 6 on Leniency. ↩︎
3.  Matthew Strathearn, Zhiqi Chen, Thomas W.Ross, “Settlements in the presence of leniency programs: Costs and benefits” Wiley online library, September 1st, 2023, https://doi.org/10.1002/mde.3991. ↩︎
4.  Website of the European Commission, “Competition Policy: Cartel cases settlement procedure”, 2023. ↩︎
5.  “Settlements in the presence of leniency programs: Costs and benefits”, supra note 3. ↩︎
6. OECD “The Future of Effective Leniency Programmes: Advancing Detection and Deterrence of Cartels”, OECD Competition Policy Roundtable Background Note, 2023. ↩︎
7.  Wouter Wils, “The European Commission’s cartel settlement procedure : An assessment after fifteen years”, Concurrences N° 3-2024, Art. N° 119526, September 2024. ↩︎
8.  OECD, “Alternatives to Leniency Programmes – Summaries of contribution”, 7 December 2023, p.10. ↩︎
9. Virgilio Pereira, “Antitrust leniency and settlement policies in the EU and the UK”, Thesis, King’s College London, September 2022, p.125. ↩︎