Exploitative Abuse After the 2026 Article 102 Guidelines: The Missing Half of the Framework

Introduction

In 2026, the European Commission adopted new Guidelines on the application of Article 102 TFEU, setting out for the first time a structured framework for assessing exclusionary abuses — conduct by a dominant undertaking that weakens competition by foreclosing rivals. The Guidelines say comparatively little about the other branch of Article 102: exploitative abuse, where a dominant undertaking harms its customers or trading partners directly, typically through unfair prices or trading conditions.

The question is whether the Commission's decision to leave exploitative abuse largely undeveloped in its Guidelines on the application of Article 102 TFEU is simply a consequence of the Guidelines' scope, or whether it leaves a gap that should now be addressed. Exploitative abuse plainly remains within Article 102 TFEU. What is less clear is what an authority is supposed to do with that category in practice. The Commission says that its principles on dominance and objective justification remain relevant to exploitative conduct, but it offers no comparable framework for deciding when an unfair price or trading condition amounts to an abuse. That becomes harder to answer where exploitation is not expressed through a price at all, particularly in digital markets involving data, privacy, quality and contractual terms.

The guidelines and the continuing place of exploitative abuse

The 2026 Guidelines on the application of Article 102 TFEU were designed to consolidate and operationalise the case law on exclusionary abuses. Exploitative conduct has never had the same settled place in Commission enforcement. The Commission's 2009 Guidance concerned exclusionary conduct only, and excessive-pricing cases have remained relatively uncommon and concentrated in particular settings, including collecting societies. OSA was decided in December 2025, only shortly before the new Guidelines were adopted, and by then the Court had clarified an important point — that an abuse under Article 102(a) TFEU need not depend on showing an unfair price at all — but it had not produced a general methodology for every form of exploitation. Timing alone does not fully explain the silence: the Commission has shown elsewhere that it can factor very recent, even pending, case law into its guidance when it chooses to, as its treatment of Booking/eTraveli in the draft Merger Assessment Guidelines illustrates, published months before the General Court's judgment in that case was handed down. The more convincing explanation is that OSA settled a threshold question without yet supplying the practical methodology the Commission would need before codifying anything.

There is also the familiar concern that Article 102 should not become a general mechanism for regulating prices. Excessive-pricing cases involve difficult questions about economic value, comparators and legitimate returns, and the Pfizer/Flynn case in the United Kingdom shows how fact-sensitive that exercise can be: the Court of Appeal's judgment treated comparators, costs, changes over time and the commercial circumstances of the parties as matters whose evidential weight has to be assessed in context. That difficulty is not, by itself, a reason to say nothing — the Commission issues guidance on plenty of other fact-sensitive tests. What Pfizer/Flynn does show is why guidance on exploitation is harder to draft than the Guidelines' exclusionary framework: even once a test exists, the benchmark itself tends to become the real dispute.

The Commission's Case AT.40394 Aspen Pharmacare investigation points in the same direction, although it should not be overstated: the Commission adopted a preliminary assessment concerning excessive prices for six off-patent cancer medicines and subsequently accepted commitments under Article 9, resolving its concerns without a finding of infringement. It would be wrong to present that outcome as proof that the Commission took commitments simply to avoid having to prove an abuse. What the case does show is that the Commission had developed a cost-plus assessment, but the proceedings ended without a contested infringement decision.

Those considerations make the Commission's restraint understandable. They do not, however, settle what paragraph 10 means in practice. Paragraph 8 says that the Commission may explain its position on questions that have not been dealt with by the case law or that remain open to interpretation, and paragraph 10 expressly identifies exploitative conduct as one such area: it confirms that the principles on dominance and objective justification carry over, without suggesting that dominance is assessed any differently depending on the theory of harm. What paragraph 10 does not do is say how an authority should decide whether a price or condition is unfair once dominance is established. The Commission could therefore have gone further, even if only by explaining the limits of what it was prepared to say. The present silence leaves national courts and competition authorities to develop much of the missing methodology themselves. The cooperation mechanisms in Regulation 1/2003 — the European Competition Network, the Commission's power to intervene in national proceedings, and the preliminary reference procedure — help keep divergent readings in check, but they operate case by case and after the fact. They are not a substitute for the methodological clarity that guidance could provide in advance, and in the meantime different national courts and authorities may reasonably reach different views on the same underlying evidential question.

From excessive prices to non-price exploitation

The case law on excessive pricing shows why that methodology matters. In United Brands — the 1978 judgment in which the Court of Justice reviewed the Commission's finding that a dominant banana supplier had charged distributors unfair prices — the Court established the familiar two-stage inquiry: first, whether the difference between cost and price is excessive; and second, whether the price is unfair in itself or by comparison with competing products. The difficulty is not really the existence of that test. It is deciding what evidence gives a reliable picture of economic value.

AKKA/LAA gave the United Brands approach considerably more detail. The Court accepted comparisons with rates in other Member States, dealt with the choice of appropriate comparators and the use of purchasing power parity adjustments, and said that a difference is appreciable where it is significant and persistent. Objective justification also remains part of the analysis. AKKA/LAA is therefore important because it shows that the economic-value inquiry does not depend on finding one perfect comparator. At the same time, the case exposes the problem that arises where the relevant value cannot easily be compared across markets. Both elements — a workable comparator, and a tolerance for differences that are genuinely significant and persistent rather than marginal — are exactly the kind of building blocks a bounded framework for exploitative abuse would need, as the final section of this post sets out.

OSA builds on that line of authority. The Court held that failing to take hotel occupancy into account may contribute to a finding of unfair pricing, but only as part of an assessment of whether the royalties are excessive in relation to the nature and scope of the use of the works and the economic value generated by that use. So the Court did give a method. The difficulty is that the method remains under-specified. How should the value generated by use actually be measured? How much weight should occupancy carry? What indicators should be used where the pricing model itself does not reflect actual use? Those are not objections to the Court's approach. They are the practical questions that remain once the principle has been stated.

The difference between the Court's judgment and Advocate General Szpunar's Opinion in OSA is useful in this respect. The Advocate General considered that occupancy was not relevant to the economic value of the works made available in hotels. The Court took a different view and treated occupancy as potentially relevant to valuation. That disagreement does not make the judgment incoherent. It does, however, show that there is still room for argument about how the valuation exercise is supposed to work in practice. OSA has moved the law forward, but it has not finished the job.

The problem becomes harder where there is no monetary price at all. Case C-252/21 Meta Platforms concerned the combination of personal data from Facebook and other sources. The Court held that a competition authority may, where necessary for the Article 102 assessment, consider whether conduct is compatible with the GDPR, while cooperating sincerely with the competent data-protection authorities. The Court also recognised that access to and use of personal data can be an important parameter of competition in digital markets. Meta was not an exploitative-pricing case, and its reasoning was not confined to exploitation. Its relevance here is narrower. It shows that Article 102 TFEU can take account of a non-price parameter where that parameter forms part of the competitive conditions in the market.

That leaves a fairly specific gap between OSA and Meta. OSA gives a valuation principle for price-based exploitation: look at the nature and scope of actual use and the economic value generated by that use. Meta shows that data and privacy can enter the Article 102 analysis, but it does not provide a general method for deciding whether a non-price condition is unfair. Assigning a conventional monetary price to something like data access or user attention is rarely realistic, and forcing the analysis into price terms would only recreate the benchmark disputes already visible in the excessive-pricing case law. A future framework would therefore have to do more than repeat that dominance and objective justification matter. It would have to explain how the relevant value or condition is to be assessed without relying on a price comparator.

Toward a bounded framework

A workable framework can start with the existing structure of Article 102 TFEU rather than creating a new test. The question is whether the condition imposed by the dominant undertaking is one that could reasonably persist under effective competition, having regard to the relevant competitive parameter and any objective justification. From there, two further points need attention.

First, where there is no market price that can usefully serve as a comparator, the Commission could look at the value extracted by the undertaking from the relevant resource. In a digital market that might mean data, user attention or contractual access. That would not mean treating all extracted value as abusive. The more difficult question is whether the value is being appropriated because the undertaking can act independently of competitive constraints, rather than because it has created value through legitimate scale, investment or innovation. A switching-based counterfactual provides one way of testing that point. The authority could ask what users would realistically have been able to obtain, supply or retain if effective competitive alternatives had constrained the dominant undertaking. This draws on reasoning already familiar from merger control and other areas of Article 102 enforcement, where counterfactual and switching analysis are used to assess competitive constraints; it is not the only way to approach the problem, and the literature is divided on how much weight switching evidence should carry outside its more familiar market-definition context, but it offers a workable starting point that keeps the assessment tied to competition on the merits. Evidence of meaningful switching to services with different terms would help show whether the disputed condition reflects competitive value. Where switching is ineffective because of network effects, lock-in or a lack of alternatives, more weight may have to be placed on the undertaking's extraction of value and on the connection between that extraction and its market power. The point is not to invent another dominance test. It is to keep the valuation exercise tied to the existing Article 102 requirement that the harm results from the exercise of a dominant position.

Second, Article 102 TFEU's relationship with other regulatory regimes needs to be stated more clearly. Compliance with the GDPR or the DMA should not automatically remove conduct from competition-law scrutiny. At the same time, a breach of another regulatory regime should not, without more, amount to an Article 102 abuse. Meta provides a useful starting point for the GDPR: competition authorities may consider GDPR compliance where it is necessary to the Article 102 assessment, but they must cooperate with the competent data-protection authorities and take account of their decisions. The same discipline should apply more generally. The competition authority still has to identify the competition-law link between the regulatory condition and the exercise of dominance.

The result would be a deliberately limited framework. The Commission would not need to produce a catalogue of every unfair condition that might arise. More useful would be guidance on the recurring evidential questions: how economic value should be assessed, when a comparator is reliable, how non-price parameters can be measured, and how conduct covered by another regulatory regime can still be connected to dominance.

Conclusion

The 2026 Guidelines are an important development in Article 102 enforcement, but they do not resolve the separate problem of exploitative abuse. Paragraph 10 is quite clear that the principles on dominance and objective justification remain relevant to exploitative conduct, and it expressly accepts that the same conduct may have both exclusionary and exploitative effects. The Court has also continued to develop Article 102(a). OSA, in particular, makes actual use and the economic value generated by that use part of the analysis of unfair pricing.

The case for further guidance is therefore narrower than a call for a general theory of fairness. What is needed is a bounded framework that keeps dominance at the centre of Article 102, treats economic value as an evidential question, and explains how non-price conditions can be examined when an ordinary price comparator is unavailable. For excessive pricing, United Brands, AKKA/LAA and OSA provide a starting point. For digital exploitation, Meta shows why privacy and data may matter to the competition assessment, while also showing why competition authorities cannot ignore the institutional role of sectoral regulators.

The Commission was entitled to focus the 2026 Guidelines on exclusionary abuse. The more difficult question is what happens after them. The case law now contains enough material for a limited framework, while paragraph 10 provides the legal bridge. Leaving the methodology entirely to national courts would mean that the Court continues to shape exploitative-abuse doctrine case by case while the Commission says little about how the pieces fit together. A modest set of principles could make that development easier to follow without trying to replace the Court's role.