This article is an opinion piece by current students or alumni of the College of Europe. The views expressed are those of the authors and do not necessarily reflect the opinions or positions of the College of Europe. Responsibility for the content lies solely with the authors.
By Giulia Galletti and Evangelia Kostopoulou
In this interview, Fernando Castillo de la Torre, Principal Legal Adviser in the Legal Service of the European Commission, where he currently leads the team for competition and mergers, discusses the evolution of EU competition litigation and enforcement, the growing importance of economic assessment in competition law, the challenges posed by digital markets and ecosystem-based competition, and how leniency and settlement procedures have reshaped the dynamics of judicial review.
Question one: Over the past two decades, EU competition enforcement has increasingly economic and effects-based. In your experience before the EU Courts, how has judicial review adapted to this shift? Do you see the Courts exercising more intense scrutiny over economic assessments, and where do you believe the limits of that review should lie?
Answer: The shift towards a more economic and effects-based approach has undeniably altered both the structure of Commission decisions and the nature of judicial review. In earlier decades, analysis under Articles 101 and 102 TFEU was often more prescriptive and category-driven. Certain types of conduct were more readily associated with infringement, and judicial review could focus on whether the practice fitted within an established legal box. The gradual move, particularly since the early 2000s, towards a contextual and economically informed assessment has required courts to engage more deeply with the factual matrix and the analytical steps underlying the Commission’s conclusions.
Judicial review has adapted in two principal ways. First, the courts are more willing to examine whether the Commission has considered all relevant circumstances, including market structure, countervailing factors, economic data, and the practical functioning of the conduct at issue. The focus is less on abstract categorisation and more on whether the decision demonstrates a coherent assessment of the competitive reality. This has led to longer, more detailed judgments, particularly in abuse-of-dominance cases, where contextual and economic circumstances are central.
Second, there has been a perceptible increase in the intensity of review in areas involving complex economic assessment. While the formal standard remains that the Commission enjoys a certain margin of assessment in matters involving technical or economic complexity, the courts now scrutinise more closely whether the methodology is robust, whether the evidentiary basis is accurate and reliable, and whether the reasoning process is internally consistent. In practice, this means that the courts will verify not only the presence of evidence but also its capacity to support the conclusions drawn. They examine whether the Commission has taken into account the relevant parameters and avoided manifest errors of assessment.
At the same time, this evolution has exposed structural tensions. Judges are not economists, and their familiarity with economic tools varies considerably. This makes the quality of the Commission’s reasoning crucial. Economic analysis must be presented in a manner that is legally intelligible and sufficiently transparent to enable effective judicial control. If the economic narrative is poorly articulated or excessively technical without explanation, it becomes vulnerable on appeal. The Legal Service’s ability to translate economic reasoning into a structured legal argument is therefore central to sustaining decisions before the courts.
Despite this increased engagement, doctrinal limits remain. The courts consistently reiterate that they are not to substitute their own economic assessment for that of the Commission where the latter has examined the relevant factors, relied on factually reliable evidence, and adopted a plausible interpretation of that evidence. The standard is not that the Commission’s conclusion must be the only possible economic reading of the facts. Rather, the evidence must be capable of supporting the conclusion reached. If multiple reasonable economic interpretations exist, the Court does not automatically replace the administrative choice with its own preferred view.
In practice, however, the line between verification and substitution is delicate. The more the law relies on multi-factor economic analysis, the greater the scope for disagreement over methodology, thresholds, and the weight attributed to particular elements. In some recent cases, particularly under Article 102, this has led to annulments where the General Court found the Commission’s reasoning insufficiently substantiated. At the same time, appellate review has occasionally corrected what was perceived as an overly demanding approach by the lower court, reaffirming the limits of judicial intervention in complex assessments.
The overall picture is therefore nuanced. Judicial review has become more searching and more visibly engaged with economic substance. Courts are less deferential in practice than they were two decades ago, particularly in the articulation of reasoning and evidentiary demands. Yet they continue to operate within a doctrinal framework that preserves a margin of administrative assessment. The balance lies in ensuring that economic analysis is rigorous, complete, and transparently reasoned, while preventing the Court from becoming a primary economic decision-maker.
Question two: The Commission acts simultaneously as investigator, prosecutor and decision-maker in competition cases. From a constitutional perspective, how do you assess the balance between administrative efficiency and procedural safeguards? Has the case law of the EU Courts meaningfully recalibrated that balance in recent years?
Answer: The concentration of investigative, prosecutorial, and decision-making functions within the European Commission has long been viewed as structurally sensitive, particularly given the magnitude of fines and the reputational consequences of infringement findings. Competition enforcement, especially in cartel and abuse of dominance cases, may exhibit quasi-criminal characteristics in terms of severity and deterrent purpose (see, for instance, the ECHR’s decision in Menarini). Nonetheless, as a matter of constitutional principle, such an administrative model is not per se incompatible with fundamental rights.
European human rights jurisprudence has consistently recognised that administrative authorities may impose sanctions, including significant financial penalties, provided that the system as a whole ensures adequate judicial protection. The decisive factor is not whether the initial decision is taken by a court, but whether the reviewing court exercises sufficiently robust control over questions of fact and law. In this sense, the constitutionality of the Commission’s dual role depends structurally on the scope and intensity of judicial review exercised by the EU Courts.
The key safeguard lies in the requirement that judicial review be capable of addressing both the legal classification of conduct and the underlying factual and evidentiary assessment. This includes the ability to examine the reliability of evidence, the coherence of reasoning, and, where relevant, the proportionality of fines. Over the past fifteen to twenty years, judicial review has become more intensive. Judgments have grown longer and more detailed, and courts engage in closer scrutiny of evidentiary elements, including economic analysis and documentary proof in cartel cases.
This evolution reflects not only a quantitative expansion of reasoning but also a qualitative shift in discourse. Courts increasingly frame their scrutiny through the language of defence rights, the presumption of innocence, and effective judicial protection under the Charter of Fundamental Rights. The narrative is no longer limited to institutional balance but explicitly tied to constitutional guarantees. At the same time, it would be inaccurate to characterise earlier case law as necessarily deferential in substance. Part of the perceived shift may stem from the increased explicitness and visibility of judicial reasoning rather than from a radical transformation of the standard itself.
Importantly, judicial control in areas involving complex economic or technical evaluations remains structured by the doctrine of margin of assessment. The courts do not exercise unlimited substitution of their own assessment for that of the Commission. Instead, they verify whether the Commission has examined all relevant elements, relied on factually reliable evidence, respected procedural rights, and provided adequate reasoning. Where these conditions are satisfied, the Commission retains a degree of discretion.
The current equilibrium may therefore be described as constitutionally conditional administrative enforcement. The institutional model remains administratively efficient, but its legitimacy depends on a form of judicial review that is sufficiently searching to safeguard defence rights without transforming the courts into primary decision-makers in complex economic matters. The recalibration that has occurred in recent years lies less in altering the system's structure and more in reinforcing the visibility and intensity of judicial guarantees within it.
Question three: Cartel and abuse of dominance cases often hinge on complex evidentiary assessments. How has the Court’s approach to standards of proof and the presumption of innocence evolved, particularly in light of the quasi-criminal nature of competition fines, and what is your view on this?
Answer: The starting point is that competition fines, especially in cartel and abuse of dominance cases, are punitive in nature and may reach extremely significant amounts. Although EU competition law is not formally classified as criminal law, the severity and deterrent character of the sanctions require guarantees comparable in substance to those applicable in criminal proceedings. This has progressively reinforced the centrality of the presumption of innocence and the requirement for rigorous proof in the case law of the EU Courts.
Over time, the presumption of innocence has increasingly gained greater value beyond cartel cases, as exemplified by the Unilever and the Servizio Elettrico Nazionale cases, in the area of abuse of dominance. It is no longer treated as a peripheral or implicit safeguard but as a structural principle governing the assessment of evidence. The Commission bears the burden of proving the infringement, and any remaining reasonable doubt must operate in favour of the undertaking. This evolution reflects the growing recognition that the quasi-criminal dimension of competition fines demands heightened procedural vigilance.
The more intricate issue, however, lies not in the formal recognition of the presumption of innocence, but in defining the object of proof. In cartel cases, the Commission must demonstrate the existence of a concurrence of wills and participation in a collusive arrangement. In abuse of dominance cases, the structure is different. Liability does not necessarily require proof of actual harm. In many instances, it is sufficient to establish that the conduct is capable of producing restrictive or exclusionary effects.
The presumption of innocence does not imply that authorities must prove with certainty that harm will materialise. Competition law frequently operates in a forward-looking and probabilistic environment. The relevant inquiry is whether, at the time of implementation, the conduct was capable of restricting competition, having regard to all relevant circumstances. Certainty about future market outcomes is rarely attainable. What must be established is a sufficiently substantiated likelihood grounded in evidence and economic logic.
The interaction between substantive standards and evidentiary thresholds is therefore crucial. As the law has moved towards a more contextual and economically nuanced approach, particularly under Article 102, the elements that must be demonstrated have multiplied. The Commission may need to show dominance, but also consider the market structure, barriers to entry, foreclosure effects, lack of objective justification, and more. The standard of proof has not necessarily been formally raised; rather, the complexity of what must be proven has increased. The evidentiary burden becomes heavier because the legal test itself is more elaborate.
This dynamic creates possible tension. On the one hand, a more economically and contextually grounded doctrine tends to enhance analytical precision and may better reflect market realities. On the other hand, it complicates enforcement and intensifies disputes over methodology and probative value. Courts must ensure that economic evidence is robust, internally consistent, and capable of supporting the Commission’s conclusions. At the same time, they must avoid transforming the requirement of proof into an unattainable standard of certainty that would undermine effective enforcement.
The adaptation of traditional evidentiary logic to digital environments illustrates this balance. In cases involving online platforms, chatrooms, or algorithmic coordination, the courts have applied established principles concerning participation and distancing. Where involvement in a collusive setting is demonstrated, liability may follow unless the undertaking shows clear and timely distancing (see, for instance, the E-Turas judgment). This does not lower the standard of proof; rather, it reflects the application of settled evidentiary doctrines to new technological contexts.
Overall, the evolution of the case law shows greater explicit reliance on the presumption of innocence and a more detailed engagement with evidentiary reasoning. The courts have reinforced the requirement that infringements be proven by sufficiently convincing evidence, while acknowledging that competition law operates in an environment of economic uncertainty. The balance lies in demanding rigour without imposing a level of certainty that is incompatible with the preventive and prospective nature of competition enforcement.
Question four: With the rise of digital markets and ecosystem-based competition, traditional analytical tools are being tested. Do you believe Article 102 TFEU is sufficiently adaptable to address digital dominance, or does the growing recourse to sector-specific regulation (such as the DMA) signal structural limits to classic antitrust enforcement?
Answer: Article 102 TFEU is, in principle, a highly flexible provision. Its open-textured language allows it to evolve across economic contexts, and it has historically accommodated new forms of conduct without the need for legislative amendment. From that perspective, there is no inherent conceptual incapacity in Article 102 to address digital dominance. The provision can, in principle, adapt to novel business models, ecosystem strategies, and platform-based practices.
The turn towards sector-specific regulation, such as the Digital Markets Act, does not therefore reflect an inability to apply Article 102. Rather, it reflects the perception that certain features of digital markets create structural and recurrent competitive concerns. Strong network effects, economies of scale and scope, data-driven advantages, and the ability of large platforms to act as gatekeepers across multiple adjacent markets generate a pattern of conduct that is not episodic but systemic. When similar concerns arise repeatedly across services and markets, relying exclusively on ex post antitrust enforcement may be seen as insufficiently agile.
The main limitation of Article 102 in this environment is operational rather than doctrinal. Enforcement under Article 102 is resource-intensive and time-consuming. It requires establishing dominance, careful market definition, often some theory of harm, and a demanding evidentiary process. Under the effects-based approach, the Commission must often conduct detailed economic analysis and consider multiple contextual elements. In fast-moving digital markets, by the time a decision is adopted and reviewed, the market may already have evolved. Where the competitive problem is structural and recurring, an ex ante regulatory framework can provide clearer, faster, and more uniform obligations.
Another important development concerns the nature of the alleged harm. Traditional Article 102 analysis has often centred on exclusionary conduct aimed at foreclosing rivals. In digital ecosystems, however, certain practices may be less about excluding competitors outright and more about extracting value from third parties operating within the ecosystem. These exploitative dynamics do not always fit neatly into classic foreclosure models. While Article 102 can in theory address exploitative abuses, articulating and proving such cases within an established doctrinal framework may be more complex.
The DMA’s focus on contestability and fairness reflects this broader shift in enforcement concerns. It sets out predefined obligations for designated gatekeepers, thereby reducing the need to establish dominance and elaborate theories of harm in each case. The relationship between Article 102 and the DMA is therefore complementary rather than substitutive. Article 102 remains available to address conduct not captured by the DMA or to pursue case-specific infringements, while the DMA addresses systemic risks inherent in certain platform structures.
The real question is not whether Article 102 is adaptable, but how far courts will go in accepting innovative applications of it in digital contexts. As enforcement theories evolve, particularly where the boundary between exclusionary and exploitative harm becomes less clear, the judicial response will determine the effective elasticity of Article 102. The provision remains conceptually capable of addressing digital dominance, but its practical reach depends on how convincingly new theories are integrated into established legal principles and how rigorously they are substantiated before the courts.
Question five: From the perspective of litigation before the EU Courts, how have leniency and settlement procedures reshaped the dynamics of judicial review? Do they strengthen enforcement credibility, or do they risk narrowing the scope for substantive judicial development?
Answer: The most significant procedural transformation has not been leniency in isolation, but the broader institutionalisation of alternative enforcement pathways: commitments under Regulation 1/2003, cartel settlement procedures, and more informal cooperation mechanisms that resemble settlement even outside the cartel context. Together, these instruments have altered the types and volumes of cases that ultimately reach the EU Courts.
Leniency, particularly since its consolidation in the early 2000s, has strengthened the Commission’s evidentiary position. Insider statements, when corroborated by documentary or other evidence, have substantially enhanced the robustness of cartel decisions. This development did not necessarily reduce cartel litigation in quantitative terms at the outset; for several years, cartel cases continued to represent a large proportion of the Court’s competition docket. What leniency primarily changed was the quality of the evidentiary record and, indirectly, the Commission’s success rate in defending its decisions on appeal.
By contrast, settlement and commitment procedures have had a more visible impact on the litigation pipeline. Settlements in cartel cases, combined with commitments, particularly in Article 102 proceedings, have significantly reduced the number of fully contested infringement decisions. Many cases are resolved administratively without reaching a decision that is litigated to final judgment. As a result, the overall number of competition cases reaching Luxembourg has diminished over time.
This evolution has two systemic consequences. First, fewer cases mean fewer opportunities for the courts to refine and develop substantive doctrine across a broad range of factual scenarios. Judicial clarification of legal standards traditionally emerges from contested litigation. When disputes are resolved through negotiated mechanisms, the scope for doctrinal elaboration inevitably narrows. In that sense, alternative procedures may reduce the breadth of jurisprudential development.
Second, the profile of the cases that reach the courts has changed. The remaining litigated cases tend to be more complex and more contentious. They often involve parties unwilling to settle because they fundamentally dispute liability, challenge the strength of the evidence, or resist structural remedies. These cases frequently raise novel legal or economic questions and may require extensive judicial engagement. Paradoxically, as case volume decreases, the intensity and depth of review in each case may increase. Fewer cases can allow the courts to devote more time and analytical attention to each dispute.
Leniency enhances evidentiary solidity, making infringement findings more resilient on appeal. Settlement reduces litigation risk and administrative burden while preserving deterrence through fines. At the same time, because settlements involve admissions of liability, the likelihood of successful challenges to the substance of the infringement is minimal. Judicial review in such cases is typically confined to specific issues, such as fine calculation or procedural aspects, rather than the core finding of infringement (see, for instance, the judgments in Printeos, Clariant and in Silgan and Crown).
Overall, the system reflects a trade-off. Enforcement efficiency and evidentiary strength are enhanced, and the Commission’s decisions are more frequently upheld. However, the reduction in fully contested judgments inevitably limits the occasions on which courts articulate and refine substantive standards in detail. The dynamic of judicial review is therefore reshaped rather than weakened: it becomes more concentrated, more selective, and more focused on structurally significant disputes rather than routine cartel litigation.
Question six: Looking ahead, do you foresee EU competition law moving towards greater integration with broader policy objectives, such as sustainability, industrial policy or strategic autonomy, or do you believe its legitimacy depends on preserving a primarily economic and consumer-welfare-oriented framework?
Answer: EU competition law is not hermetically sealed off from broader Union objectives, but its integration with them is necessarily mediated through a competition-based rationale. Competition has always been conceived as a foundational component of the internal market and thus as a value that supports wider economic and social goals. It is not an end in isolation. At the same time, competition law cannot become a general instrument for the direct pursuit of environmental, industrial, or geopolitical objectives. Its legitimacy depends on remaining anchored in protecting the competitive process.
Intervention under Articles 101 and 102 TFEU must therefore be framed in terms of competitive structure, market dynamics, or competitive outcomes. Conduct cannot be condemned solely because it appears environmentally undesirable or politically problematic. The legal analysis must demonstrate a distortion of competition, a restriction of market access, or an adverse impact on competitive parameters. Broader concerns may inform the context, but they cannot replace the competition narrative.
That said, the way competition is conceptualised may evolve. If competition is understood not merely as price rivalry but as the preservation of pluralism, openness, and contestability, particularly in digital and information markets, then safeguarding competitive structures may indirectly support democratic values or media plurality. In such cases, the doctrinal anchor remains competition, even if the societal implications extend beyond a narrow consumer-welfare metric.
Broader objectives may also enter through justification frameworks. Under Article 102, objective necessity or objective justification may, in principle, accommodate certain public-interest considerations, provided they are substantiated by credible evidence and proportionate to the aim pursued. Similarly, under Article 101, the case law has recognised structured spaces where non-competition considerations are integrated, such as in the context of professional regulation, sports governance, or specific social-policy arrangements. These are not open-ended gateways; they are carefully circumscribed doctrines requiring rigorous proof and a clear connection between the conduct and the asserted legitimate objective (see, for instance, the Google Android and Albany cases).
In practice, the constraint is often evidentiary rather than theoretical. Undertakings invoking sustainability, public health, or other public-interest rationales must demonstrate that the conduct genuinely pursued those aims at the time of implementation. Courts are unlikely to accept ex post rationalisations unsupported by contemporaneous documentation or consistent strategic reasoning. The threshold for accepting such justifications remains demanding.
The direction of travel, therefore, appears to be one of cautious integration rather than transformation. Broader policy objectives may shape the interpretation of competitive harm or inform the assessment of justification, but they do not displace the need for a disciplined, competition-based analysis. The system’s credibility and operability depend on maintaining doctrinal clarity and avoiding conflating competition enforcement with general regulatory policymaking.
Fernando CASTILLO DE LA TORRE
Fernando Castillo de la Torre is Director (Principal Legal Adviser) in the Legal Service of the European Commission, where he now leads the team for competition and mergers. He has been working in the Legal Service since 1996 in the teams for internal market, competition, external relations, and agriculture, fisheries and animal and plant health. He studied both law and political science in Madrid (Spain), graduating in 1989 and 1990 respectively, and undertook postgraduate studies in EU law at the College of Europe in Bruges (Belgium).
He worked in the chambers of the President of the European Court of Justice between 1997 and 2002. He has been agent for the European Commission in almost 500 cases in the European Court of Justice, and has pleaded in many of the most important competition cases over the last 20 years. He has published widely on the relationship between legal orders (international, EU and national), judicial review and litigation before EU Courts, competition law, the internal market and the external relations of the EU.
Giulia GALLETTI
Giulia Galletti is an LL.M. candidate in European Legal Studies at the College of Europe. She holds a Master's degree in Law from the University of Bologna and an LL.M. from King’s College London, and has been admitted to the Italian Bar.
She previously trained at an international law firm and completed a Blue Book traineeship at the European Commission (DG COMP). Her research interests include Competition Law and Economics, Financial Services, and International Trade.
She currently serves as President of the Competition Society at the College of Europe.
Evangelia KOSTOPOULOU
Evangelia Kostopoulou is an LL.M. candidate in European Legal Studies at the College of Europe. She holds a master's degree in law and technology from Tilburg University and is admitted to the Greek Bar. She previously trained at EY Law, one of the largest national law firms in Greece.
Her academic interests lie in EU Competition Law, and she is particularly interested in questions of consumer protection, sustainability transitions and digital platform governance. Alongside these substantive concerns, she has a strong interest in competition procedure and the procedural safeguards that structure EU antitrust enforcement.
She currently serves as a Board Member of the Competition Society at the College of Europe.
