Anticompetitive Agreements Within Apple’s Distribution Network and Abuse of a Situation of Economic Dependency: End of an Antitrust Saga

On May 13, 2026, the French High Court (Cour de cassation) upheld the reasoning of the Paris Court of Appeal (Dec. n° 20/08582, Oct. 6, 2022) in its entirety in a ruling dismissing the appeals (Cour de cassation, May 13, 2026, n° 22-22.623, 22-22.647, 22-22.677, 22-22.727). The judges thus confirmed the existence of a cartel in the distribution of Apple products, with wholesalers Ingram Micro and Tech Data, as well as Apple’s abuse of economic dependence towards its premium resellers, putting an end to this antitrust saga that started in 2020.

As a reminder, the French competition authority (“FCA”) had received a complaint in 2012 from a distributor of specialized high-end Apple products, eBizcuss (Cour de cassation, May 13, 2026, para. 3), company which went into bankruptcy that same year (E.g., para. 9), regarding anticompetitive practices in Apple’s distribution network. After its investigation, the FCA fined Apple €1,1 billion, Ingram Micro €62,9 million and Tech Data €76,1 million on March 16, 2020 (FCA, dec. n° 20-D-04, March 16, 2020; FCA, Press release, March 16, 2020).

Three anticompetitive practices were targeted by the FCA in this decision. First, the allocation of products and customers between Apple and its two wholesalers Tech Data and Ingram Micro. Secondly, the imposition of selling prices by Apple on its premium resellers and thirdly, the abuse of Apple’s premium resellers situation of economic dependency by treating them discriminatorily, provoking supply difficulties and unstable remuneration conditions (Cour de cassation, May 13, 2026, para. 4; FCA, Press release, March 16, 2020).

The Paris Court of Appeal had significantly reduced the fines imposed in a ruling issued on October 6^th, 2022 (Paris Court of Appeal, dec. n° 20/08582, Oct. 6, 2022) for three reasons. First, the proof of imposition of resale prices was insufficient and this practice was hence not characterized. Second, the anticompetitive agreement regarding division of consumers and products had been sanctioned for a duration of eight years (2005 to 2013). However, according to the judges, the evidence of an agreement was only sufficient from 2009 and thus the duration of the anticompetitive agreement was only four years and was less damaging to the economy. Thirdly, the rates imposed by the FCA, in the calculus of the fines, were too high and did not respect the principle of proportionality of the sanction (for further explanations: Main Developments in Competition Law and Policy 2022 - France | Kluwer Competition Law Blog).

Apple, Ingram Micro, Tech Data, as well as the President of the FCA filed an appeal to the French High Court, which were joined in the decision from May 13^th, 2026.

The FCA President held that the anticompetitive agreement started in 2005, using email exchange on December 2, 2005, between Apple and wholesalers stating that there was an “allocation problem” (Cour de cassation, May 13, 2026, para. 30), the process being unreliable and too hefty. The email from Apple stated “More and more troubleshooting for our ACs regarding wholesalers!!!! Unable to allocate products to accounts (iPods on Acropom/Actimac - PowerBooks on France Syst)”, meaning wholesalers were not free to decide which reseller customers they supplied (E.g., para. 30). However, this email was ambiguous at its best and was not evidence of an anticompetitive agreement, the Paris Court of Appeal having freely determined the effective date based on sufficient evidence (E.g., para. 31-34), which was 2009 and not later contrary to what Apple, Ingram Micro and Tech Data were arguing (E.g., para. 35-41).

The FCA President also challenged the non-characterization of the imposition of resale prices, saying that Apple had indirectly encouraged mandated prices on its wholesalers through a series of measures such as transparent communication of the prices charged in its own channels, inciting to adopt the same prices, monitoring and control over promotions practiced by authorized resellers (E.g., para. 97), etc. However, the judges upheld the reasoning of the Paris Court of Appeal, the evidence only being circumstantial (E.g., para. 98-105). For instance, there was no direct evidence to establish the company communicated with all its distributors to ask them to maintain a designed price level (E.g., para. 98). Furthermore, a significant number of Premium Resellers acknowledged that they were able to run promotional campaigns or that Apple provided no guidelines regarding the monitoring of advertised prices (E.g., para. 102). The resellers thus had a certain autonomy and flexibility, and there were no prices imposed by Apple.

On these two points, the High Court simply upheld the reasoning of the Paris Court of Appeals, not bringing much more to the table. However, the appeals filed by Apple, Ingram Micro and Tech Data are a bit more of interest. Indeed, even if the judges upheld the lower court’s ruling, it offers some interesting insights into few concepts.

Right to a fair trial and administrative procedure

The companies argued that there had been a violation of the adversarial principle, equality of arms, and procedural fairness. In their view, the use of exhibits appended to the FCA reports before the Court of Appeals, as well as unlisted exhibits or documents from competitors protected by trade secret, had deprived them of an effective hearing prior to their appeal, and thus of their right to adversarial proceedings (E.g., para. 11 & 14-15).

Cour de cassation clarified these procedural issues and dismissed the appeals (E.g., para. 12 & 16-29). First, regarding the documents in the appendix, they had indeed been brought to the parties’ attention, and the parties therefore had an opportunity to discuss them, even if joined only in the appendix (E.g., para. 12). On the other hand, the unlisted seized documents had all been provided in the form of a complete copy of the digital sealed file and subsequently in their listed version during the appeal filed with the Court of Appeal (E.g., para. 18). Finally, the FCA had reminded the companies of their option to waive trade secret protection to gain access to the documents, which they had not done (E.g., para. 19). The companies would therefore have had the opportunity to discuss these documents, had they exercised their rights. Nevertheless, the Court clarified the matter, emphasizing that there is no violation of the adversarial principle unless it can be established that elements were derived from them that altered the nature of the offense being prosecuted, the seriousness of the facts, or the duration of participation in that offense, and that, in doing so, the situation of the companies involved would have been affected (E.g., para. 17). This was not the case here since the FCA did not exceed grounds for appeal by sticking to the anticompetitive practices targeted by its decision. Notably, the allocation of products and customers between Apple’s wholesalers was characterized as a restriction by object.

Customer restriction and restriction by object under competition law

Apple and its wholesalers argued that there was a lack of evidence on several points: first, regarding proof of customer allocation, which requires a body of serious, specific, and consistent evidence in the absence of direct proof (E.g., para. 22-24). However, there were indeed numerous documents and pieces of evidence, including direct proof of this practice, such as emails explicitly referring to the implementation of the disputed allocation system, or the existence of Apple’s interference in the sales process, which limited supply options, demonstrating the coercive nature of this process and the resellers’ compliance with this practice, as they went along with it (E.g., para. 25-29).

Second, a violation of Article 4(b) of the Vertical Block Exemption Regulation was alleged (E.g., para. 42-43). According to the appellants, a customer restriction — which deprives the vertical agreement of the benefit of the exemption — can only be identified where it appears that the distributor or wholesaler is, because of the disputed prohibition, exposed to a loss of market share or customers. In this regard, there was a fluctuation in the market shares of wholesalers regarding certain Apple Premium Resellers over certain periods (E.g., para. 43). The High Court methodically dismissed this appeal (E.g., para. 44-57). Indeed, the restriction may take form of direct or indirect obligations aimed at dissuading the distributor from reselling to its customers (E.g., para. 48). The judges then demonstrate Apple’s interference, which was intended to allocate customers exclusively between the two competing authorized wholesalers and thus maintain their market shares within a stable range, without regard for the normal play of competition, resulting in the preservation of customers for Apple (E.g., para. 49-52). This closed system allocated customers between the only two authorized wholesalers, constituting a restriction on customers. It was hence irrelevant to show that wholesalers would have suffered the loss of customers to which this system inherently exposed them (E.g., para. 54).

Finally, it was argued that the restriction by object should not have been upheld according to the companies (E.g., para. 58-60), as it must be interpreted strictly and may be applied only to certain types of coordination among firms that demonstrate a sufficient degree of harmfulness (E.g., para. 58). Cour de cassation rejected this claim as well in a very lengthy, detailed and educational explanation, mobilizing European case law (E.g., para. 61-84).

It first reminds that certain from of collusions may be so harmful to competition that an examination of their effects is unnecessary (quoting: ECJ, European Superleague Company, December 21, 2023, C-333/21, para. 166; Servier e.a./Commission, June 27, 2024, C-201/19, para. 76: Cour de cassation, May 13, 2026, para. 61). To determine this degree of harm, the competition authorities examine three criteria: the content of the agreement, the objective pursued, and the economic and legal context in which it operates (quoting: ECJ, CB/Commission, September 11, 2014, C-67/13 P, para. 57; ECJ, Budapest Bank e.a., April 2, 2020, C-228/18, para. 67; Cour de cassation, May 13, 2026, para. 62-63). Indeed, restriction by object must be proven in all circumstances and cannot be presumed, even in the presence of a clear restriction, such as the setting of minimum resale prices (Cour de cassation, June 28, 2023, n° 21-26.015; see: Interbank settlement : The French Supreme Court rejects the appeal lodged by the Competition Authority and puts an end to the saga of interbank fees for the exchange of cheque images | Concurrences) or a customer restriction, as in this case. Even though horizontal restrictions, among direct competitors, are more harmful than vertical restraints, such vertical restraints can become a restriction by object if the harm is sufficiently significant (quoting: ECJ, Allianz Hungária Biztosító e.a., March 14, 2013, C-32-11, para. 43; Cour de cassation, May 13, 2026, para. 64), even in the presence of pro-competitive effects (quoting: ECJ, Energias de Portugal e.a., October 26, 2023, C-331/21, para. 104; Cour de cassation, May 13, 2026, para. 64) or a legitimate objective (quoting: ECJ, Budapest Bank e.a., April 2, 2020, C-228/18, para. 52; Cour de cassation, May 13, 2026, para. 66), provided that the harm outweighs these factors.

The judges then examined cautiously the three criteria. Regarding content of the agreement, the judges drew on the facts set forth throughout the decision. While it was a vertical restriction, it still had a significant impact. The agreement between Apple, Tech Data and Ingram Micro established a system of monitoring and control over wholesalers, depriving them of freedom in the choice of customers and allocated products. They were intended solely to allocate customers and products exclusively between the two competing wholesalers that were members of the distribution network (Cour de cassation, May 13, 2026, para. 67-70). Thus, there was sufficient and reliable evidence to conclude that this type of agreement was, by its very nature, particularly harmful to the proper functioning of competition (E.g., para. 70).

Regarding the objectives pursued, Apple and its wholesalers cited pressure from multi-brand sellers in the upstream market for the distribution of IT and consumer electronics products. The allocation system would have been intended to preserve the competitive advantage that wholesalers derived from being certified sellers of the distribution network (E.g., para. 71). However, the shortage of Apple products was orchestrated by the company itself. Furthermore, the allocation system was maintained beyond the alleged periods of constraint. Finally, the implementation of a new distribution system starting in 2013 – the date the anticompetitive agreement ended – demonstrated that less anti-competitive means existed to address those issues (E.g., para. 71). The legitimate objective was therefore not established and did not negate the particularly harmful nature of the disputed practice (E.g., para. 72).

Finally, regarding economic context (E.g., para. 73-76), paragraph 73 is of particular interest. It describes how Apple’s market power is far superior to its market shares (16.3% in volume and 24.7% in value) due to Apple’s specific closed ecosystem. Indeed, Apple has differentiated products for many reasons. First, Apple-branded products are innovative and high-end, exhibiting greater complementarity with one another than competing products, in that their respective hardware and software features are designed to function optimally when combined. Conversely, their interoperability with competing hardware is limited, which increases switching costs and explains partly why Apple is one of the high-tech companies with the most brand-loyal consumers. While inter-brand competition exists on the upstream market, it is significantly reduced by the brand’s reputation, this limited interoperability with other brands (E.g., para. 73), and the fact that only two wholesalers are authorized to distribute in the wholesale market (E.g., para. 74). There should thus be vigilance to maintain fair competition as these conditions already reduce competitive pressure. This was not the case with the anticompetitive agreements which were, by their very nature, likely to affect the normal play of competition (E.g., para. 76), characterizing a restriction by object.

Furthermore, these conditions make Apple an unavoidable economic partner for distributors of high-end electronic products, playing a role in the characterization of an abuse of economic dependence in this case.

Abuse of economic dependence in French competition law

This French provision (article L. 420-2, 2° of the Code de commerce) penalizes any abuse by a supplier whose distributors are in a position of economic dependence, or conversely, by a distributor whose suppliers are in a position of economic dependence.

Regarding the supplier, Apple, in this case, this dependence is assessed based on several factors: the supplier’s share of the distributors’ revenue, its share in the market, as well as the supplier’s market reputation. Finally, French case law also requires a demonstration that no equivalent solution exists for distributors, meaning that it is impossible to obtain, within a reasonable timeframe, a solution that is technically and economically equivalent to the contractual relationship it has established with that supplier (E.g., para. 86).

This last condition is difficult to meet, particularly in the mass retail sector (as illustrated by the Cora case for instance: Cour de Cassation, December 10, 1996, n° 94-16.192), where there are several suppliers and thus equivalent alternatives, without considering the specific circumstances of the case. Apple therefore argued that there was a sufficient equivalent alternative — Samsung — whose market shares were greater than or equal to Apple’s (Cour de cassation, May 13, 2026, para. 85), which could have gone either way in front of the High Court regarding case law. However, this claim was also rejected, marking a victory for the enforcement of abuse of economic dependence.

Firstly, Apple’s market reputation and its share in distributors’ revenue were not hard to prove, with market shares fluctuating between 78 and 90% (E.g., para. 87). Then, it was stated that Samsung did not constitute an alternative supplier. While market share was comparable, the retail networks were not. Apple is the only company to operate a network of specialized retailers in physical stores dedicated to its own brand, with strong recognition and consumer’s attachment to the brand’s unique environment, making it a sole key player (E.g., para. 88). Furthermore, because of these specific characteristics, and the significant costs associated with reconfiguring the store layout, Apple Premium Resellers could not switch status to general reseller status during their contract (E.g., para. 89). To conclude, none of the alternative solutions could be qualified of being equivalent (E.g., para. 91) as they were not substitutable to Apple’s products. Hence, Apple Premium Resellers were in a situation of economic dependency towards Apple (E.g., para. 90).

Apple should have ensured that these Premium Resellers were not subjected to measures that any rational business would refuse, and that they accept only because of this state of dependence (E.g., para. 92). However, this was not the case, as abusive conducts were put into place by Apple, such as discriminatory supply terms, or the opening of Apple Stores in competition with certain Premium Resellers within their catchment areas (E.g., para. 93-94). Those abuses had effects on the market, creating a competitive disadvantage for these resellers. The three criteria for abuse of economic dependence (an abuse, a position of economic dependence and an effect on the market) were thus met, and the sanction was justified, for a remarkable application of article L. 420-2, 2° of the Code de commerce, that will hopefully not be the last since digital ecosystems prone to reinforce power imbalances and create situation of dependency.