Пульс · Документы · Европейский союз · Суд Европейского Союза (CJEU) — InfoCuria

Федеральный верховный суд ЕС: злоупотребление доминирующим положением Google на рынке онлайн-рекламы

Opinion of Advocate General Kokott delivered on 12 November 2026.

судебное решение 2026-11-12 150 882 знаков Онлайн-реклама Поисковые системы
Скачать На сайте источника Файл оригинала
Действующая редакция. Последнее изменение зафиксировано 2026-11-13.

OPINION OF ADVOCATE GENERAL

KOKOTT delivered on 12 November 2026 ( 1 )

Case C ‑ 826/24 P

European Commission v Google LLC, Alphabet Inc.

( Appeal – Competition – Abuse of dominant position – Market for online search advertising intermediation in the EEA – Decision finding an infringement of Article 102 TFEU and of Article 54 of the EEA Agreement – Exclusive supply obligations – Contractual restrictions – Exclusivity clause – Placement clause – Prior authorisation clause – Exclusionary abuse – Intel criteria – Duration of contractual commitment – Single and continuous infringement )

Table of contents

I. Introduction

II. Background to the dispute

A. The contested contractual practice of Google

B. Administrative procedure and contested decision

C. Action for annulment and judgment under appeal

1. Pleas in law and subject matter of the appeal

2. Second plea in law regarding the exclusivity clause

3. Third plea in law regarding the placement clause

4. Fourth plea in law regarding the prior authorisation clause

D. Procedure before the Court of Justice and forms of order sought

III. Assessment

A. Main complaint in the first, in the second and in the third grounds of appeal: error of law in applying the Intel criteria

B. Scope and weight of the Intel criteria in the context of the overall assessment of restrictive effects on competition

1. Generally applicable criteria for assessing the existence of restrictive effects on competition in connection with Articles 101 and 102 TFEU

2. Specific characteristics in connection with Article 102 TFEU

C. Assessment of the application of the Intel criteria to the clauses at issue in the judgment under appeal

1. First Intel criterion: extent of the dominant position

2. Second and fourth Intel criteria: market coverage rate and strategy aimed at excluding as-efficient competitors

(a) Preliminary remark

(b) Exclusivity clause

(c) Placement clause

(d) Prior authorisation clause

(e) Interim conclusion

3. Third Intel criterion: duration of the GSAs in particular

(a) Examination of the duration of contractual commitments in the contested decision

(1) Exclusivity clause

(2) Placement clause

(3) Prior authorisation clause

(b) Examination of consideration of duration by the General Court

(1) Exclusivity clause

(2) Placement clause

(3) Prior authorisation clause

(c) Conclusions

D. Grounds of appeal in detail

1. First ground of appeal regarding the exclusivity clause

(a) First branch: complete assessment of the Intel criteria and of the relevant circumstances of the case

(b) Second branch: sufficient demonstration of exclusionary effects

(c) Third branch: inter alia undue reversal of the burden of proof to the detriment of the Commission with regard to the assessment of the duration of the exclusivity clause

2. Second ground of appeal regarding the placement clause

(a) First branch: complete assessment of the Intel criteria and of the circumstances of the case

(b) Second branch: sufficient demonstration of exclusionary effects

(c) Third branch: inter alia undue reversal of the burden of proof to the detriment of the Commission with regard to the assessment of the duration of the placement clause

3. Third ground of appeal regarding the prior authorisation clause

(a) First branch: complete assessment of the Intel criteria and of the circumstances of the case

(b) Second branch: sufficient demonstration of exclusionary effects

(c) Third branch: inter alia undue reversal of the burden of proof to the detriment of the Commission with regard to the assessment of the duration of the prior authorisation clause

4. Interim conclusion

5. Fourth ground of appeal: single and continuous infringement

E. Referral back to the General Court

IV. Conclusion

I. Introduction

1. The European Commission has brought the present appeal against a judgment of the General Court ( 2 ) (‘the judgment under appeal’) by which it annulled a Commission decision (imposing a fine) addressed to the respondents ( 3 ) (‘the contested decision’) on grounds of an inadequate examination of the relevant circumstances of the case and thus failure to demonstrate infringements of Article 102 TFEU.

2. The appeal raises fundamental questions in connection with the demonstration of infringements of Article 102 TFEU which are based on contractual commitments and produce restrictive effects on competition. Since the underlying (coordinated) practices may also fall within the scope of Article 101 TFEU, regard must be had, first, to the need to interpret Articles 101 and 102 TFEU consistently. ( 4 ) Second, against that background, it is necessary to clarify the scope of the Commission’s obligations with regard to examination and demonstration in relation to such effects as well as the associated standard of proof.

3. More precisely, it concerns the requirements for demonstrating actual or potential exclusionary effects to the detriment of competitors arising from contractual commitments which an undertaking with a dominant position on (multi-sided) digital markets imposes on its customers, in this case website publishers. The main focus is the different criteria which the Court of Justice recognised in the judgment in Intel v Commission ( 5 ) (‘the Intel I judgment’) for examining and demonstrating the existence of such exclusionary effects (‘the Intel criteria’). These include the duration of the contractual commitment, ( 6 ) which is the central point in contention in this dispute. In the judgment under appeal, the General Court criticised the Commission for examining that criterion only inadequately with regard to the underlying individual contracts and for thus failing to demonstrate the existence of exclusionary effects which, all together , are harmful to competition. That is disputed by the Commission in its appeal, in essence, on the ground that the General Court erred in law in assessing the criterion of duration and its relationship with the other Intel criteria, the fulfilment of which it largely considered to be proven.

4. The Court of Justice thus has an opportunity to clarify the scope of the Intel criteria, in particular the criterion of the duration of contractual commitments and its relationship with the other criteria, and to shed further light on the associated requirements with which the Commission must comply in examining and demonstrating exclusionary effects. In that regard, consideration must also be given to the main objection raised by the respondents, based on competition on the merits, with which they were successful at first instance. According to that objection, the Commission failed, in view of the relatively short duration of the commitment in the individual contracts, to demonstrate exclusionary effects because, after they had expired, the volume of demand covered by the contracts up to then could have been contested by competitors, but they have been unsuccessful in doing so because their offering was allegedly of inferior quality.

II. Background to the dispute

A. The contested contractual practice of Google

5. The facts underlying the legal dispute are, in essence, as follows. ( 7 )

6. Google LLC is a company established in Mountain View, California (United States), which is active worldwide – including within the European Economic Area (EEA) – in the information and communications technology sector and has specialised in online products and services. It is the subsidiary of Alphabet, Inc., which is also established there (together, ‘Google’ or ‘the respondents’) and is known in particular for its search engine and its general search service ‘Google Search’. That service allows internet users to find and access, with their browser of choice and by means of hyperlinks, websites meeting their needs.

7. A key element of Google’s business model is its flagship service, its digital platform Google Search, and the associated services. More precisely, its business model is based on the interaction between, on the one hand, online products and services, which are generally offered to (internet) users free of charge, and, on the other, online advertising services, ( 8 ) from which it derives the vast majority of its revenues. Those revenues are generated principally through the Google Search service. Google’s business model is aimed primarily at increasing the number of users of Google Search in order to make the search service more attractive for online advertising and to be able to sell more advertising services to third-party providers. By collecting data from internet user interactions, Google also continuously improves its ability to present relevant search results and advertisements. In so doing, as an important digital platform, it benefits from network effects and other advantages typical of such platforms. ( 9 )

8. The Google Search service, accessible at the address ‘www.google.com’ or at similar addresses with a specific national extension, yields search results appearing on internet users’ screens. Those results are selected by the search engine either according to general criteria or according to a specialised logic, without the publishers of the sites to which they link needing to pay a fee to Google.

9. However, Google Search’s results pages also display websites linked to payments from their publishers. Those results, commonly called ‘advertisements’ or ‘ads’, also follow a search carried out by the internet user. They are, however, distinguished from unremunerated results of general or specialised search queries by the additional designation ‘ad’ or ‘sponsored’. Their preferential display and their respective ranking on the results pages of Google Search are linked to payment commitments from advertisers undertaken in online auctions. These are managed via Google’s digital auction platform AdWords . ( 10 )

10. Since 2003, Google has also operated a digital online advertising intermediation platform called AdSense . In that regard, Google has developed various related services including an online search advertising intermediation service called AdSense for Search (AFS). ( 11 ) That intermediation service enables publishers of third-party websites, which are independent of Google and whose websites contain integrated search engines, to display online search ads from Google Search when users carry out searches on their websites. In this way, the revenues generated from the display of those ads can be shared between intermediaries and publishers.

11. In order to use the AFS intermediation services, advertisers had to associate their ads with keywords which users of the websites concerned were expected to use in carrying out an online search. In order to determine which ads were to be displayed (and how) in response to an online search, Google primarily took into account, on the one hand, the price that the advertiser had indicated that it was willing to pay in an auction held for that purpose and, on the other hand, the relevance of those ads to the search query and, thus, the probability that the user would click on those ads. In principle, the advertiser paid the price resulting from the display of its ad only where the user actually clicked on it.

12. To use AFS, website publishers concluded inter alia two types of contracts with Google.

13. First, publishers could conclude, for one or more of their websites, an ‘online contract’, that is, a standard-form, non-negotiable contract. Google categorised publishers that had concluded such a contract as ‘online partners’. That kind of contract is not the subject of the present proceedings.

14. Second, publishers could conclude, for one or more of their websites, a ‘Google Services Agreement’ (GSA). Unlike the online contracts, GSAs were negotiated with each website publisher individually. Google categorised publishers that had opted to conclude a GSA as ‘direct partners’. Irrespective of this, Google drew up template GSAs, which were amended on a number of occasions – notably in March 2009 – and were frequently used as a basis by contractual partners. Moreover, in order to conclude a GSA, direct partners were required to complete an order form. In that form they had to specify whether they wished to use AFS or a different AdSense service to list the websites covered as well as their addresses.

15. Up until March 2009, the template GSA contained inter alia two clauses. The clause known as the exclusivity clause stipulated that the direct partner was not to implement on the websites listed in the order form a service which was identical to or substantially similar to the services supplied by Google under the GSA or which was in direct competition with those services. The clause known as the English clause stipulated that, subject to the exclusivity clause, the direct partner and Google had to endeavour to reach agreement on a new order form before approaching another online search or advertising service provider. Moreover, in the event that the direct partner and Google did not manage to reach agreement on a new order form and the direct partner decided to approach such a provider, that clause provided that Google could make an offer matching the terms proposed by that provider.

16. From March 2009, the template GSA contained two new clauses which replaced the exclusivity clause and the English clause. Under the clause known as the placement clause , the direct partner was to display a minimum number of Google ads on websites using AFS. In addition, it was not permitted to display competing ads from other intermediaries above those from Google or directly adjacent to them. The clause known as the prior authorisation clause required that partner to obtain Google’s approval before every change to the display of online search ads on a listed website, including competing ads. Google could refuse to give its approval only for certain reasons. It was assumed that Google had given its approval if it did not respond within 15 business days.

17. All GSAs containing the prior authorisation clause also included the placement clause. However, not all GSAs with the placement clause necessarily also contained the prior authorisation clause.

18. According to the order form corresponding to the March 2009 template GSA, screenshots of the search results pages of the websites listed in that order form for the use of AFS (‘mock-ups’) were to be included. The mock-ups had to illustrate the number, format and placement of Google ads on those pages.

B. Administrative procedure and contested decision

19. In January 2010, the Bundeskartellamt (Federal Cartel Office, Germany) transferred to the Commission a complaint that had been lodged by Ciao GmbH against Google.

20. On 30 November 2010, the Commission initiated proceedings against Google pursuant to Article 2(1) of Regulation (EC) No 773/2004. ( 12 )

21. On 31 March 2011, 30 March 2012 and 30 January 2013, respectively, Microsoft Corporation, Expedia Inc. and Initiative for a Competitive Online Marketplace lodged complaints against Google.

22. On 13 March 2013, the Commission adopted a preliminary assessment within the meaning of Article 9 of Regulation (EC) No 1/2003, ( 13 ) which related, inter alia, to the contractual obligations imposed by Google in respect of online search advertising.

23. On 3 April 2013, 21 October 2013 and 31 January 2014, Google offered the Commission commitments to address the concerns set out in the latter’s preliminary assessment.

24. On 16 May 2014 and 2 July 2015, Deutsche Telekom AG, among others, lodged a complaint against Google.

25. On 21 April 2016, Microsoft and Ciao withdrew their complaints.

26. On 14 July 2016, the Commission decided to initiate the proceedings provided for in Article 2(1) of Regulation No 773/2004 with respect to the exclusivity, placement and prior authorisation clauses that were provided for by GSAs (together, ‘the clauses at issue’). On the same day, it adopted a statement of objections pursuant to Article 10 of that regulation. In that statement, it informed Google that the clauses at issue were liable to constitute an abuse of a dominant position and therefore infringe Article 102 TFEU.

27. On 9 September 2016, Google informed the Commission that it had sent letters to all direct partners notifying them that it was removing the exclusivity and prior authorisation clauses in their entirety as well as certain provisions of the placement clause.

28. On 20 March 2019, the Commission adopted the contested decision.

29. In that decision, it found that from 2006 to 2016 Google had committed three separate infringements, constituting, together, a single and continuous infringement of Article 102 TFEU. Accordingly, it ordered Google to pay a fine of EUR 1 494 459 000, part of which jointly and severally with Alphabet Inc.

30. The Commission found that in that period Google had held dominant positions in 30 out of 31 relevant national markets for online search advertising (excluding the Portuguese market) and in the distinct market for online search advertising intermediation in the EEA.

31. The Commission calculated Google’s underlying market shares, first, on the basis of its gross and its net revenues and, second, on the basis of the number of online search queries (the latter only for the national markets for online search advertising). In addition, from 2006 to 2016, Google had faced only limited competition from other online search advertising providers or intermediaries for such advertising, including Microsoft and Yahoo!. There were numerous barriers to entry and expansion, such as the need for significant investments in order to establish and maintain such platforms, network effects and the fact that nearly all advertisers were tied to Google’s AdWords digital auction platform. Lastly, there was a lack of countervailing buyer power on the part of advertisers and website publishers. ( 14 )

32. According to the Commission, the exclusivity clause had constituted, from 1 January 2006 until 31 March 2016, an abuse of a dominant position in so far as that clause was contained in GSAs which ‘typically’ included all the websites of direct partners displaying online search ads. ( 15 ) Primarily, it considered that that clause imposed an exclusive supply obligation on direct partners, which, as such, was contrary to Article 102 TFEU in accordance with the judgment in Hoffmann-La Roche v Commission ( 16 ) (‘the judgment in Hoffmann-La Roche ’), without it having been necessary to verify that it had been capable of restricting competition in the light of all the circumstances of the case. ( 17 ) The Commission considered, in the alternative, that that clause constituted such an infringement because it had been capable of restricting competition. ( 18 ) It categorised as ‘all sites direct partners’ direct partners that had typically included all their websites in at least one of their GSAs. The exclusivity clause had (i) deterred those partners from sourcing from Google’s competing intermediaries, (ii) prevented those intermediaries from accessing a significant part of the market for online search advertising intermediation, (iii) possibly deterred innovation, (iv) helped Google to maintain and strengthen its dominant position on the national markets for online search advertising in the EEA (excluding the Portuguese market), and (v) possibly harmed consumers. In addition, the Commission found that the English clause had exacerbated the capability of the exclusivity clause to restrict competition. The Commission rejected the objective justifications put forward by Google. ( 19 )

33. According to the Commission, from 31 March 2009 until 6 September 2016, the placement clause had constituted an abuse of a dominant position on the ground that it was capable of restricting competition in the light of all the circumstances of the case and Google had not demonstrated that that clause had been objectively justified. ( 20 ) The placement clause had reserved the most prominent space on the websites covered by it for Google ads. Direct partners had been required under that clause to display, in the most prominent space on those websites, a ‘block’ of three ‘wide format’ Google ads when the search was made on a desktop device and at least one Google ad when the search was made on a mobile device. The underlying analysis of the (potentially) restrictive effects on competition of the placement clause essentially corresponds, in terms of its structure and its conclusion, to the analysis of the exclusivity clause. In addition, the Commission found that the binding nature of the mock-ups had exacerbated the capability of the placement clause to restrict competition. The Commission rejected the objective justifications put forward by Google. ( 21 )

34. Lastly, the Commission found that, from 31 March 2009 until 6 September 2016, the prior authorisation clause had also constituted an abuse of a dominant position. In the light of all the circumstances of the case, it had been capable of restricting competition and Google had not demonstrated its objective justification. ( 22 ) The underlying analysis, including the examination of justifications, ( 23 ) corresponds, in terms of its structure and its conclusion, to the analyses of the exclusivity and placement clauses.

35. The Commission characterised the three abovementioned abuses of a dominant position together as a single and continuous infringement of Article 102 TFEU, which had lasted from 1 January 2006 until 6 September 2016. By the clauses at issue, Google had pursued the same objective, namely to exclude its competing intermediaries in order to maintain and strengthen its position both on the online search advertising intermediation market and on the online search advertising markets themselves and, consequently, its position on the general search market. In addition, those clauses were complementary in that they sought to deter direct partners from sourcing competing search ads and to prevent Google’s competing intermediaries from accessing a significant part of the market for online search advertising intermediation in the EEA. ( 24 )

C. Action for annulment and judgment under appeal

1. Pleas in law and subject matter of the appeal

36. Google based the action for annulment which it brought against the contested decision before the General Court on five pleas in law.

37. By the first plea in law, Google claimed that the Commission had made an erroneous definition of the market for online search advertising intermediation and, consequently, had not demonstrated that Google had a dominant position on that market. The General Court rejected that plea in law as unfounded. ( 25 )

38. By the second, third and fourth pleas in law, Google complained that the clauses at issue did not constitute an abuse within the meaning of Article 102 TFEU. The General Court ultimately upheld those pleas in law and therefore annulled the contested decision to that extent.

39. The fifth plea in law was directed against the fine imposed by the Commission. The General Court did not make any findings on that plea in law in the light of the annulment of the contested decision on the basis of the second, third and fourth pleas in law. ( 26 )

40. The assessment of the second, third and fourth pleas in law in the judgment under appeal forms the main subject matter of this appeal and will be presented in more detail below. In the absence of a cross-appeal by Google, that does not apply to the first and fifth pleas in law.

2. Second plea in law regarding the exclusivity clause

41. The second plea in law concerning the assessment of the exclusivity clause consisted of three parts. First, that clause did not constitute an exclusive supply obligation within the meaning of the judgment in Hoffmann-La Roche . Second, the Commission had therefore been required to examine whether that clause was capable of restricting competition. Third, the Commission had not demonstrated in the contested decision that that was the case. The General Court ultimately upheld all three parts of that plea in law.

42. With regard to the first and second parts of the second plea in law, the General Court found, having regard to the judgment in Intel I and the judgment in Unilever Italia Mkt. Operations , ( 27 ) that, in view of Google’s challenge during the administrative procedure, the Commission was unable to rely on the judgment in Hoffmann-La Roche , but was required to assess specifically, taking into account all the relevant circumstances of the case, whether the clauses at issue, including the exclusivity clause, were capable of restricting competition, which it had done in the contested decision only in the alternative. ( 28 ) The Commission does not challenge those findings by the General Court in its appeal.

43. The General Court ultimately also held that the third part of the second plea in law concerning the capability of the exclusivity clause to restrict competition was well founded and upheld it in order to annul the contested decision to that extent. ( 29 )

44. The General Court did, in the context of the third part of the second plea in law, reject as unfounded several complaints raised by Google, in particular with regard to the deterrent effect of the exclusivity clause on all sites direct partners established in the contested decision and their being prevented from accessing a significant part of the market. ( 30 ) In the context of the assessment of the duration for which all sites direct partners were obliged, by virtue of that clause, to source all or most of their requirements from Google, however, the Commission had failed to take into consideration all the relevant circumstances of the case.

45. In addition, according to the General Court, the Commission had not established that the exclusivity clause in the GSAs concluded with those partners could have produced a foreclosure effect, owing to its market coverage, between 1 January and 31 March 2016. ( 31 ) That finding does not form part of the appeal lodged by the Commission and is thus final.

46. The General Court held that those two ‘errors’ vitiated all the restrictions of competition identified in the contested decision. The Commission had not therefore demonstrated to the requisite legal standard that the exclusivity clause had been capable of deterring direct partners from sourcing from Google’s competing intermediaries or preventing those intermediaries from accessing a significant part of the market for online search advertising intermediation in the EEA and thus of having the assumed foreclosure effect. Nor had the Commission demonstrated to the requisite legal standard that that clause possibly deterred innovation, helped Google to maintain and strengthen its dominant position on the national markets for online search advertising at issue and possibly harmed consumers. That also applied to the exacerbating effect of the English clause which had been established. ( 32 )

3. Third plea in law regarding the placement clause

47. In the context of the assessment of the third plea in law concerning the placement clause , which consisted of two parts, the General Court rejected the first part regarding the determination of the scope of that clause as unfounded. ( 33 ) The Commission had not established that direct partners were thereby prohibited from displaying competing ads below Google ads. ( 34 ) Google, for its part, had not proved, however, that in practice the click-through rates of competing ads could have been at least comparable with that of its own ads, as displayed in accordance with that clause. The Commission had therefore been correct to find that the placement clause reserved the most prominent spaces of the results pages of direct partners for Google ads. ( 35 )

48. The General Court held, on the other hand, that the second part of the third plea in law was well founded, upholding it and annulling the contested decision to that extent. ( 36 )

49. The Commission had, in its examination of the effects of the placement clause and the associated restrictions of competition, also taken into consideration its duration and market coverage rate. ( 37 ) However, it had failed to take into consideration all the relevant circumstances of the case in the context of the assessment of the duration for which direct partners had been obliged to reserve the most prominent spaces of their results pages for Google ads by virtue of the placement clause. The General Court reached that conclusion even though, in assessing the deterrent effect of that clause established in the contested decision and the resulting impossibility for Google’s competing intermediaries to access a significant part of the market for online search advertising intermediation, it rejected the majority of Google’s complaints as unfounded. ( 38 )

50. In addition, according to the General Court, the Commission had not established that the placement clause could have produced a foreclosure effect, owing to its coverage, between 1 January and 6 September 2016. ( 39 ) That finding – like that regarding the exclusivity clause – does not form part of the appeal lodged by the Commission and is thus final.

51. The General Court held that those two ‘errors’ vitiated all of the restrictions of competition identified in the contested decision. The Commission had not therefore demonstrated to the requisite legal standard that the placement clause had been capable of deterring direct partners from sourcing from Google’s competing intermediaries or preventing them from accessing a significant part of the market for online search advertising intermediation in the EEA and, consequently, of having the foreclosure effect found in that decision. Nor had it demonstrated to the requisite legal standard that that clause had possibly deterred innovation, helped Google to maintain and strengthen its dominant position on the national markets for online search advertising at issue and possibly harmed consumers. ( 40 ) Lastly, the Commission had not demonstrated that the binding nature of the mock-ups had exacerbated that foreclosure effect of the placement clause or that the mock-ups as such were capable of having that effect. Accordingly, the contested decision had to be annulled to the extent that it found that the placement clause constituted an infringement of Article 102 TFEU, without it being necessary to rule on the other arguments of Google under the third plea in law. ( 41 )

4. Fourth plea in law regarding the prior authorisation clause

52. The fourth plea in law concerning the prior authorisation clause likewise consisted of two parts.

53. In the context of the first part , Google had essentially argued that that clause did not produce foreclosure effects, had not helped it to maintain or strengthen its dominant position on the national markets for online search advertising at issue and had neither deterred innovation nor harmed consumers. ( 42 ) The General Court ultimately considered that first part to be well founded and annulled the contested decision to that extent, without examining the second part regarding objective justification. ( 43 )

54. As grounds, similarly to its assessment of the second and third pleas in law, the General Court held in essence that the Commission had taken into account duration and market coverage rate when it examined the effects of the prior authorisation clause in the context of the restrictions of competition identified in the contested decision. However, it failed to take into consideration all the relevant circumstances of the case in the context of the assessment of the duration for which direct partners had been obliged, by virtue of that clause, to request prior authorisation in writing from Google before changing the display of competing ads on their results pages. The General Court reached that conclusion even though, in the context of the assessment of the deterrent effect of that clause vis-à-vis direct partners established in the contested decision, the resulting impossibility for Google’s competing intermediaries to access a significant part of the market for online search advertising intermediation and the as-efficient competitor test, it rejected nearly all of Google’s complaints as unfounded. ( 44 )

55. In addition, according to the General Court, the Commission had not established that the prior authorisation clause could have produced a foreclosure effect, owing to its market coverage, between 1 January and 6 September 2016. ( 45 ) That finding – like those regarding the exclusivity and placement clauses – does not form part of the appeal lodged by the Commission and is thus final.

56. The General Court held that those two ‘errors’ vitiated all the restrictions of competition identified in the contested decision. The Commission had not therefore demonstrated to the requisite legal standard that the prior authorisation clause had been capable of deterring direct partners from sourcing from Google’s competing intermediaries or preventing those intermediaries from accessing a significant part of the market for online search advertising intermediation in the EEA and, consequently, of having the foreclosure effect which had been found. Nor had it demonstrated to the requisite legal standard that that clause had possibly deterred innovation, helped Google to maintain and strengthen its dominant position on the national markets for online search advertising at issue and, lastly, possibly harmed consumers. The first part of the fourth plea in law therefore had to be upheld, without it being necessary to rule on the other arguments of Google under that plea. In addition, the contested decision had to be annulled to the extent that it found that the prior authorisation clause constituted an infringement of Article 102 TFEU. ( 46 )

D. Procedure before the Court of Justice and forms of order sought

57. By document lodged at the Registry of the Court of Justice on 4 December 2024, the Commission brought an appeal against the judgment under appeal.

58. By document lodged at the Registry of the Court of Justice on 27 February 2025, the respondents applied, in respect of the interveners at first instance, Surfboard Holding BV and Vinden.NL BV, for confidential treatment of certain information included in their response and in Annexes R. 2, R. 3 and R. 4 thereto, as had been granted by the General Court at first instance. By order of 22 May 2025, ( 47 ) rectified by order of 30 June 2025, ( 48 ) the President of the Court of Justice granted that application.

59. The Commission claims that the Court should:

– set aside the judgment under appeal;

– dismiss the action in so far as the respondents challenge the capability of the clauses at issue to restrict competition between 31 March 2009 and 31 December 2015;

– order the respondents to pay the costs of the appeal proceedings and of the proceedings before the General Court; or

– in the alternative, refer the case back to the General Court and reserve the costs.

60. The respondents contend that the Court should:

– dismiss the appeal; and

– order the Commission to pay the costs; or

– in the alternative, refer the case back to the General Court and reserve the costs.

61. At the hearing on 15 July 2026, the parties presented oral argument and answered the questions put by the Court of Justice.

III. Assessment

A. Main complaint in the first, in the second and in the third grounds of appeal: error of law in applying the Intel criteria

62. By its first three grounds of appeal, the Commission challenges certain parts of the General Court’s assessment of the first, second and third pleas in law concerning the exclusionary effects arising from the clauses at issue in so far as they relate to the period from 2006 to 2015. In so far as the General Court annulled the contested decision on grounds of failure to demonstrate exclusionary effects in 2016, that is not contested by the Commission. Accordingly, the judgment under appeal is final in that respect. The fourth ground of appeal concerns the question whether it follows from the successive use of those clauses – as established in the contested decision – that there is a single and continuous infringement of Article 102 TFEU.

63. As was explained in point 41 et seq. above, the General Court’s assessment of whether the clauses at issue were capable of producing an exclusionary effect and restricting competition, which led it to find ‘errors’ on the part of the Commission, is similar both in terms of its structure and its content. In all three instances, the General Court criticises the Commission for failing to take into consideration and to assess sufficiently the relevant circumstances of the case, in particular in connection with the respective duration of the contracts containing those clauses, in order to demonstrate (actual or at least potential) exclusionary effects and restrictions of competition. ( 49 )

64. The Commission challenges that finding by its first, second and third grounds of appeal. Those grounds of appeal are essentially based on the complaint that in its assessment the General Court misunderstood the scope of the criteria for establishing exclusionary effects ( 50 ) recognised by the Court of Justice in the judgment in Intel I and their weighting in relation to one another. In particular, the General Court attached too great importance to the criterion of the duration of the commitments in the individual contracts and imposed on the Commission excessive requirements to demonstrate something when considering and examining that criterion and unduly reversed the burden of proof to its detriment in that respect. The appeal does not therefore seek an (inadmissible) review of the General Court’s assessment of the facts and evidence in that regard or a ‘rewriting’ of the contested decision, as Google repeatedly maintains, but alleges an error of law in the context of the assessment of the relevant criteria for demonstrating exclusionary abuse.

65. In fact, in the judgment in Intel I , the Court of Justice developed for the first time – departing from its judgment in Hoffmann-La Roche – additional evidentiary requirements for demonstrating abuse by the Commission in relation to exclusivity commitments such as loyalty rebate schemes. Those requirements relate to the case where the dominant undertaking concerned submits, during the administrative procedure, on the basis of supporting evidence, that its conduct was not capable of restricting competition and, in particular, of producing the alleged foreclosure effects. In that case, the Commission must analyse, above all, the extent of the dominant position of the undertaking concerned on the relevant market (first Intel criterion), the share of the market covered by the challenged practice (second Intel criterion), its conditions, arrangements and duration (third Intel criterion) and the possible existence of a strategy aiming to exclude competitors that are at least as efficient as the dominant undertaking from the market (fourth Intel criterion). ( 51 )

66. According to the Court of Justice, the Intel criteria are – as Google correctly asserts – applicable to contractual exclusivity commitments like those in the clauses at issue. ( 52 ) As the General Court also found ( 53 ) and is not contested by the Commission in its appeal, it is no longer sufficient in demonstrating abuse to establish in accordance with the finding made in the judgment in Hoffmann-La Roche that the dominant undertaking imposes unilaterally or contractually on customers the obligation to source all or most of their requirements exclusively from it, in order to tie them to itself. ( 54 )

67. The Court of Justice has subsequently clarified that obligation with regard to the examination and demonstration to the effect that the Commission is required, in establishing (exclusionary) abuse within the meaning of Article 102 TFEU, to examine all the relevant circumstances of the case. ( 55 ) In the judgment under appeal, however, the General Court inferred from this – incorrectly, as I will show below – that that obligation with regard to examination not only covers the overall assessment of the different Intel criteria, as such, including the duration of the contractual commitment, but also requires a particularly detailed analysis of all the relevant circumstances in respect of each individual criterion, failure to comply with which may, in itself, result in the annulment of the contested decision. ( 56 ) In my view, the General Court thereby imposed excessive requirements as to the scope of the examination of the criterion of duration and did not sufficiently relate its conclusion to the assessment of the other Intel criteria or weigh it in the context of those criteria.

68. In the contested decision, although the Commission relied principally on the judgment in Hoffmann-La Roche in order to establish the exclusionary effects of the exclusivity clause, it undoubtedly also analysed those effects in the alternative by reference to the Intel criteria. ( 57 ) It did the same for the other two clauses at issue. ( 58 ) As the General Court recognised in principle, ( 59 ) the Commission did take into consideration and examine the duration of the clauses in a large number of GSAs as a relevant circumstance, but not all GSAs in the detailed and comprehensive manner required by the General Court (see also points 44, 49 and 54 above). ( 60 )

69. I will begin by considering, in the context of the first three grounds of appeal, the scope and the weight of the Intel criteria, in particular the criterion of the duration of the contractual commitment, in connection with the necessary overall assessment of restrictive effects on competition and the Commission’s related obligations with regard to examination and demonstration (see section B). I will then examine the central complaint in all those grounds of appeal, which claims that, in assessing whether the Commission had satisfied the Intel criteria in the contested decision, the General Court did not comply with the legal framework laid down by those criteria (see section C). Lastly, I will briefly address, in the light of the result of that analysis, the individual complaints raised in the first, second and third grounds of appeal, including in so far as they have not yet been answered by the preceding analysis (see section D).

B. Scope and weight of the Intel criteria in the context of the overall assessment of restrictive effects on competition

70. In examining restrictive effects on competition, regard should be had, first, to the criteria generally applicable to infringements of Articles 101 and 102 TFEU and, second, to the specific criteria which apply only to abuse of a dominant position within the meaning of Article 102 TFEU.

1. Generally applicable criteria for assessing the existence of restrictive effects on competition in connection with Articles 101 and 102 TFEU

71. According to settled case-law, a finding of the existence of actually or at least potentially restrictive effects on competition of unilateral or coordinated practices within the meaning of Article 101 or Article 102 TFEU requires an overall assessment of all the relevant circumstances of the case. That applies not only in connection with the concept of ‘restriction of competition by effect’ – arising from a coordinated practice – within the meaning of Article 101(1) TFEU, ( 61 ) but also to unilateral actions, in particular in the form of abusive conduct, at issue in the present case, intended to exclude or impede under Article 102 TFEU. Such a practice must also, unless it is in any case presumed to be particularly harmful, ( 62 ) at the very least, be capable of restricting competition (on the merits) or – in other words – of having the actual or potential effect of giving rise to such a restriction. ( 63 ) The relevant circumstances of the case to be examined relate to the relevant overall context, that is to say, the conduct in question itself, the market(s) in question, the functioning of competition on that or those market(s) and the characteristics of the sector concerned as a whole. ( 64 )

72. The requirements for demonstrating restrictive effects on competition are not any different, in principle, within the framework of the respective scopes of Articles 101 and 102 TFEU. In both cases the question is whether (potential) competitors can still access or remain on the market in question or improve their market position as a result of the practice complained of. This is linked to the question of the resulting increase in barriers to market entry or market foreclosure effects. A consistent interpretation in that regard is all the more necessary because the practice in question may have to be examined on the basis of the conditions of both Article 101 TFEU and Article 102 TFEU, albeit taking into account the specific characteristics of those provisions. ( 65 ) That is also true in the present case as the Commission could also have reviewed the contractual commitments, namely the GSAs containing the clauses at issue underlying the alleged infringements of Article 102 TFEU, in terms of their compatibility with Article 101 TFEU. ( 66 )

73. The criteria to be applied in examining an individual case, in the context of an overall assessment, in order to establish a market foreclosure effect have been developed in longstanding case-law. They essentially correspond to the Intel criteria recognised for exclusionary abuse inter alia in connection with exclusivity commitments. Thus, as Google mentioned at the hearing, the Court of Justice examined in the judgment in Delimitis the existence of a restriction of competition by effect in the form of foreclosure and its attribution to its authors, inter alia, on the basis of the following criteria: ( 67 ) the existence of a bundle of similar supply contracts (or a network of contracts), which tie the customers to the producers concerned and impede access to the market by competing producers or cumulatively deny such access; the quantity of supplies covered; the conditions under which competition operates on the relevant market, that is, the number and the size of producers present on the market, the degree of saturation of the market and customer fidelity to existing brands; the (appreciable) extent or significance of the contribution of the supply contract at issue and its contracting parties to the foreclosure in the light of the market share held by producer, the number of customers tied to it in relation to the total number of outlets available and the duration of the contract. In that regard, the Court of Justice made clear that even a relatively short duration of a contractual tie may make a significant contribution to the sealing-off of the market if the producer holds a relatively strong market position. ( 68 ) The Commission adopted those requirements in its Guidelines on vertical restraints. ( 69 )

74. The Intel criteria are clearly linked to the abovementioned criteria. Thus, the first Intel criterion, the extent of the dominant position of the undertaking concerned to be considered, is based in particular on its share on the relevant market, which is an important indicator of its market power. ( 70 ) The second Intel criterion, the market coverage rate , corresponds to the criterion of contractually agreed supplies or the volume of demand covered by similar supply contracts. The third Intel criterion concerning conditions, arrangements and duration is similar to the conditions under which competition operates on the relevant market, in particular on the basis of the contractual arrangements designed and implemented for that purpose. Lastly, the fourth Intel criterion, the existence of a possible strategy aiming to exclude competitors , is linked to criteria for attributing an (at least potential) market foreclosure to the contracting parties.

75. The General Court also found in the judgment under appeal, with reference to paragraph 36 of the Commission’s Priorities Communication, ( 71 ) that, in general, the longer the duration of the exclusive supply obligation, the greater the likely foreclosure effect, it being specified that if the dominant undertaking is an unavoidable trading partner for all or most customers, even a short duration can lead to anticompetitive foreclosure. ( 72 )

76. The Commission points out in that communication, correctly in my view, that the switching of supplier by customers may be rendered difficult due to the duration of the exclusive purchasing obligation, a longer duration of the obligation meaning a greater foreclosure effect. That statement cannot, however, be understood to mean that in the case of an exclusive purchasing obligation of short duration, anticompetitive foreclosure can exist only if the dominant undertaking is an unavoidable trading partner for all or most customers. In such a case, a foreclosure is only all the more likely. ( 73 ) Ultimately, however, such an understanding, towards which the General Court appears to incline, ( 74 ) is also immaterial because the Priorities Communication is not binding on the Court of Justice. ( 75 ) Rather, it must be considered, in accordance with existing case-law, that the duration of the exclusive supply obligation must be evaluated in an overall assessment in conjunction with the other criteria, particular importance being attached to the market position of the supplier and the market coverage rate. ( 76 )

77. Lastly, as the Commission correctly submits, the need for an overall assessment taking into consideration all the relevant factors and evidence follows from the case-law on the concept of a significant impediment to effective competition ( 77 ) in merger control law. ( 78 ) If complaints are raised against the Commission’s assessment in that regard before the General Court, according to that case-law, it must ‘assess whether all the relevant factors and evidence on which the Commission has relied and which may be regarded as established, including those which have not been disputed, are sufficient to demonstrate the existence of a significant impediment to effective competition’. According to the Court of Justice, the General Court therefore errs in law if, in response to the complaints raised by the applicant, it ‘fail[s] to carry out … an overall assessment of the relevant factors and findings, in order to ascertain whether the Commission had demonstrated the existence of a significant impediment to effective competition’. ( 79 )

2. Specific characteristics in connection with Article 102 TFEU

78. According to the case-law of the Court of Justice, the specific characteristics to be taken into consideration in connection with Article 102 TFEU in examining the existence of restrictive effects on competition (see point 72 above) include the question whether the conduct complained of is capable of excluding as-efficient competitors from the market in question. The criteria for determining efficiency are the cost structure, capacity to innovate or quality of the dominant undertaking. ( 80 )

79. It is true that the Court of Justice has pointed out that the as-efficient competitor test , which is applicable primarily to price competition , ( 81 ) may be inappropriate in the case of non-pricing practices, such as a refusal to supply, or significant barriers to entry and is only one of a number of methods for assessing whether a practice is capable of producing exclusionary effects. ( 82 ) However, in examining whether contractual exclusivity commitments like the clauses at issue here are capable of producing exclusionary effects, it has – ambiguously, in my view ( 83 ) – not ruled out the use of that test entirely. ( 84 )

80. The Court of Justice has, in turn, subsequently made clear that, in order to establish abuse of a dominant position, it is necessary, as a rule , to demonstrate that the conduct complained of deviates from the means of competition on the merits and thereby results – actually or potentially – in equally efficient competing undertakings being excluded or impeded and thus in a restriction of competition on the market concerned or related or neighbouring markets. ( 85 ) It must be considered in that regard that those two constitutive elements of the concept of abuse are, in principle, to be examined and established separately. ( 86 ) However, abuse can also be established on the basis of the fact that that conduct of the dominant undertaking has ‘the actual or potential effect – or even the object – of impeding potentially competing undertakings at an earlier stage, through the placing of obstacles to entry or the use of other blocking measures or other means different from those which govern competition on the merits , from even entering that or those market(s) …’. ( 87 )

81. I understand that case-law to mean, in any event, that the placing of barriers to entry or the use of other blocking measures by the dominant undertaking is sufficient, in principle, to attribute to it the resulting exclusionary effects to the detriment of competitors and to find it at fault for abusive conduct, without it being necessary in each case to establish additionally a deviation from competition on the merits. ( 88 )

82. In doing so, the Court of Justice, first, draws the correct conclusions from the particular responsibility of the undertaking in a dominant position, emphasised in settled case-law, for maintenance of competition on the markets in question ( 89 ) and the necessary market structure to that end, in particular given that the degree of competition there is already weakened precisely because of the presence of that undertaking. ( 90 ) That applies a fortiori to the structure of digital markets, like those at issue here. That structure involves an ecosystem characterised by significant barriers and network effects, which contribute to making the entry, maintenance or emergence of even a competitor as efficient as the undertaking with a dominant position practically impossible. ( 91 )

83. Second, that understanding is consistent with the case-law concerning contractual exclusivity commitments within the meaning of Article 101 TFEU, according to which a market foreclosure effect may be indicated by a particularly strong market position of a contracting party, even where the duration of contractual commitments is relatively short (see points 73 to 76 above).

C. Assessment of the application of the Intel criteria to the clauses at issue in the judgment under appeal

84. I will now consider whether the General Court complied with the legal framework recognised in the case-law of the Court of Justice, as described in point 71 et seq. above.

85. In the judgment under appeal, the General Court examined all the Intel criteria and the manner in which the Commission took them into consideration in the contested decision. In doing so, it essentially confirmed – except with regard to the duration of the GSAs – that the Commission’s assessment and findings were free from errors of law and errors of assessment. Because the establishment of exclusionary effects is the result of an overall assessment of those criteria, I will also take into consideration below the examination of those criteria in the judgment under appeal which are not, as such, called into question in the appeal either by the Commission or – in the absence of a cross-appeal – by the respondents. For the sake of simplicity, I will analyse the clauses at issue together in so far as they are the subject of that assessment and examination. That appears to be all the more expedient since the General Court made similar findings in that regard. I will have regard to the Intel criteria in the manner in which compliance with them by the Commission was examined by the General Court.

1. First Intel criterion: extent of the dominant position

86. With regard to the extent of Google’s dominant position in the majority of the national markets at issue for online search advertising and in the market for online search advertising intermediation, the General Court confirmed the Commission’s findings in full. ( 92 ) According to those findings, Google’s market share had increased on those markets between 2006 and 2016 and almost no competitors remained in 2016. Furthermore, those markets were characterised by the existence of significant barriers to entry and expansion and by a lack of countervailing buyer power on the part of advertisers and publishers. In particular, the scale and network effects had rendered the emergence of new competitors difficult. ( 93 ) That assessment, among other factors, led the General Court to the preliminary conclusion that the exclusivity clause in GSAs with all sites direct partners – subject to the examination of its duration – could have been capable of producing the foreclosure effect accepted in the contested decision. ( 94 ) The same applies, in principle, to the GSAs containing placement ( 95 ) and prior authorisation clauses. ( 96 )

87. I would point out, additionally, that, according to the now final findings in the contested decision, ( 97 ) Google’s dominant position in the abovementioned markets was based on particularly high market shares, namely of well above [ confidential ] % on most national markets for online search advertising and of [ confidential ] % on the market for online search advertising intermediation in the EEA in 2016. ( 98 ) In the latter case, the Commission found that there had been a – particularly significant – sharp increase since 2006 (with a market share already exceeding [ confidential ] %). ( 99 ) As was also found by the General Court, that is all the more important as those markets were characterised, first, by the existence of significant barriers to entry and expansion and scale and network effects and, second, by a lack of competitors and countervailing buyer power. In my view, those findings are, in themselves, strong indications that the clauses at issue – irrespective of short contractual commitments in some cases – were not only capable of producing exclusionary effects, but actually did so. The General Court took the same view in principle, but then qualified those circumstances again with regard to the criterion of duration (see point 103 et seq. below).

2. Second and fourth Intel criteria: market coverage rate and strategy aimed at excluding as-efficient competitors

(a) Preliminary remark

88. The General Court examined the criterion of market coverage by the clauses at issue in each instance in connection with the question of the impossibility for Google’s competing intermediaries to access a significant part of the market for online search advertising intermediation and endorsed the contested decision in that regard. ( 100 ) Before that, it considered and confirmed the – closely related – effect of each of those clauses in deterring direct partners, that is, customers of its intermediation services, from approaching Google’s competing intermediaries. ( 101 )

89. After addressing the criterion of market coverage, the General Court rejected Google’s complaints that the Commission incorrectly failed to take into account the as-efficient competitor test which excluded the finding of exclusionary effects by the clauses at issue. In view of the wording of those clauses, their market coverage and the structure and functioning of the markets in question (Google’s high market shares, barriers to entry and expansion, including network effects) and the insufficient evidence submitted by Google, ( 102 ) that test was not applicable. Furthermore, it is irrelevant whether Google pursued a strategy of excluding competitors as competitive as it (competing intermediaries) from the market. ( 103 ) Rather, the Commission could merely demonstrate those exclusionary effects by relying on several other relevant elements. ( 104 ) According to the General Court, the satisfaction of the as-efficient competitor test was therefore irrelevant for the purposes of establishing a restriction of competition by the clauses at issue.

90. On account of the close connection between those different aspects in demonstrating whether the clauses at issue were capable of producing exclusionary effects and restricting competition, I will examine them – and, in that regard, the first, second and third grounds of appeal – together in more detail below. This appears to me all the more necessary since the General Court essentially rejected Google’s complaints in that regard and thus confirmed the findings in the contested decision.

(b) Exclusivity clause

91. The General Court first confirmed the effect of the exclusivity clause, established in the contested decision, in deterring all sites direct partners from sourcing at least part of their requirements from other intermediaries in order to display competing ads on their websites or on certain of their associated pages. ( 105 ) It equally confirmed that that clause was capable of preventing Google’s competitors from accessing a significant part of the market for online search advertising intermediation in the EEA. ( 106 ) In that regard, it also essentially rejected Google’s complaints regarding purportedly insufficient market coverage ( 107 ) and the purported failure to examine the as-efficient competitor test. ( 108 )

92. In that connection, the General Court again referred to the increase in Google’s market share between 2006 and 2016, the lack of competitors and countervailing buyer power and the existence of significant barriers to entry and expansion, including effects of scale and network effects. ( 109 ) Furthermore, after an in-depth examination, ( 110 ) it made the preliminary finding – subject to an examination of the duration of the contractual commitments – that, owing to its market coverage, the exclusivity clause in GSAs with all sites direct partners could have been capable of producing the foreclosure effect accepted in the contested decision. That did not apply, however, to market coverage in the period between 1 January and 31 March 2016. ( 111 )

(c) Placement clause

93. As regards the deterrent effect of the placement clause on direct partners, the General Court essentially reached the same conclusion, ( 112 ) especially since that clause reserved for ads from Google the most prominent spaces on the results pages of those partners and thus the spaces generating the highest click-through rates. ( 113 )

94. The Commission was able to consider that the placement clause was capable of restricting competition by deterring those partners from sourcing their requirements from Google’s competing intermediaries. ( 114 ) The General Court essentially rejected as unfounded the complaints raised by Google against that finding. ( 115 ) Thus, the inclusion of the placement clause in the GSAs indicated that, notwithstanding the allegedly superior quality of AFS, Google considered that, absent that clause, direct partners would have had a commercial interest in sourcing from other intermediaries. ( 116 ) In that regard too, the General Court therefore arrived at the preliminary conclusion – subject to the examination of the duration of the contractual commitments – that the placement clause could have deterred certain of those direct partners from sourcing at least part of their requirements from Google’s competing intermediaries. ( 117 )

95. In that connection, the General Court referred again to the increase in Google’s market share between 2006 and 2016 and the structure and functioning of the markets in question (lack of competitors and countervailing buyer power, significant barriers to entry and expansion and scale and network effects) ( 118 ) in order to confirm – subject to the examination of duration – that the placement clause could have been capable of producing the foreclosure effect accepted in the contested decision. ( 119 ) The General Court also rejected as unfounded several complaints raised by Google – except for 2016 – with regard to the impossibility, established in that decision, for Google’s competing intermediaries to access a significant part of the market for online search advertising intermediation as well as market coverage by that clause. ( 120 ) That also applies to the complaints alleging that the Commission, first, had not sufficiently examined the as-efficient competitor test and, second, had failed to demonstrate that an as-efficient competitor could have emerged on the market for online search advertising intermediation or was capable of being excluded from that market. ( 121 )

96. Subject to the examination of duration, the General Court found, lastly, that the Commission was correct to consider that (having regard to the coverage by the exclusivity clause in GSAs by all sites direct partners) the coverage by the placement clause could be sufficient to enable it to be capable of producing a foreclosure effect between 31 March 2009 and 31 December 2015. ( 122 ) That did not apply, however, to the establishment of such an effect owing to coverage between 1 January and 6 September 2016. ( 123 )

(d) Prior authorisation clause

97. The General Court’s assessment regarding the effect of the prior authorisation clause of deterring direct partners from sourcing from competing intermediaries and regarding the impossibility for the latter, on account of market coverage in particular, to access a significant part of the market for online search advertising intermediation overlaps with or is similar to the corresponding assessment regarding the other two clauses. ( 124 ) That assessment relates inter alia to the interaction with the placement clause, the deterrent effect of which could be exacerbated by the prior authorisation clause. ( 125 ) In that regard, the General Court also took into consideration the extent of Google’s dominant position on the market for online search advertising intermediation and particularly the significant barriers to entry and expansion there. ( 126 ) In that connection, the General Court rejected all Google complaints as unfounded. ( 127 ) It therefore reached the preliminary conclusion – subject to the examination of duration – that the prior authorisation clause could have been capable of producing the foreclosure effect found in the contested decision. ( 128 )

98. Furthermore, the General Court essentially confirmed the finding made in the contested decision that between 2009 and 2015 (but not for 2016) – inter alia in combination with the market coverage by the placement clause – the prior authorisation clause had covered a significant part of the market for online search advertising intermediation in the EEA. As regards the impossibility, established in that decision, for competing intermediaries to access a significant part of the market for online search advertising intermediation, the General Court rejected several complaints raised by Google as unfounded. ( 129 ) That also held for the complaints regarding the inadequate examination of the as-efficient competitor test and the alleged absence of a strategy aiming to exclude such competitors (see also points 89 and 96 above). The Commission could merely demonstrate the capability of the prior authorisation clause to produce a foreclosure effect by relying on several relevant elements, without relying on that test. Those elements included the substance of the prior authorisation clause, the fact that that clause, together with the exclusivity clause and the placement clause, covered a significant part of the online search advertising intermediation market and the structure and functioning of that market (extent of Google’s dominant position, above all on the basis of its very high market shares and barriers to entry and expansion, in particular network effects). ( 130 )

99. The General Court concluded from this, first, that, subject to an assessment of duration, the Commission was correct to consider that – having regard to the market coverage by the exclusivity clause and the placement clause – the market coverage by the prior authorisation clause could be sufficient to enable that clause to be capable of producing a foreclosure effect between 31 March 2009 and 31 December 2015. Second, Google was not justified in arguing that the Commission had failed to demonstrate that a competitor as efficient as it could not emerge on the market for online search advertising intermediation or was capable of being excluded from that market. ( 131 )

(e) Interim conclusion

100. This detailed examination of the effects of the clauses at issue on direct partners and competitors, based inter alia on their market coverage, permits no other conclusion than that the General Court also considered that those clauses – particularly in combination – had been capable, in principle, of producing the exclusionary effects found in the contested decision and that the Commission had complied with its obligation as to demonstration in that regard. However, the General Court reached that conclusion for each of those clauses subject to the proviso that the Commission should have taken into consideration and examined sufficiently the duration of the GSAs and the existence of a unilateral termination right. ( 132 ) As I will show below, I consider that to constitute an error of law.

3. Third Intel criterion: duration of the GSAs in particular

101. I will therefore examine, as a next step, whether the General Court was able, in view of the Commission’s findings which it confirmed, regarding the capability of the clauses at issue to produce exclusionary effects and thus to restrict competition, to attach such great weight to the (relatively short) duration of individual contractual commitments in order to call into question the conclusion it had reached up to that point. In my view, that is not the case. Even if it were correct that in the contested decision the Commission did not examine and assess the criterion of duration – which it did indeed take into consideration – with the precision required by the General Court in respect of all the individual contracts, the General Court is, for its part, at fault for failing to relate its finding sufficiently to the result of its examination of the other Intel criteria and to weigh it, as part of an overall assessment, in order to justify its conclusion that the decision should be annulled. The General Court was, however, obliged to do so on the basis of its function as the court responsible for the assessment of facts and the need to review fully the relevant facts and evidence on which the Commission relied in the contested decision in establishing an infringement under Article 102 TFEU. ( 133 )

102. First of all, however, it is necessary to consider in greater detail the content of the contested decision in that regard. The aim is not a reassessment of the relevant facts and evidence, which is inadmissible at the appeal stage, or even a substitution of the reasoning in the contested decision, but it is necessary in order to lay the groundwork for the examination of whether the General Court erred in law in assessing its content on the basis of the Intel criteria.

(a) Examination of the duration of contractual commitments in the contested decision

103. In general and in response to the objections already raised by Google in the administrative procedure, which are based partly on the allegedly short duration of the contractual commitments and the unilateral termination rights of the contracting parties, the Commission made the following findings in the contested decision regarding each of the clauses at issue.

(1) Exclusivity clause

104. With regard to the exclusivity clause, the Commission assessed inter alia the duration provided for in the GSAs concerned ( 134 ) as part of its analysis of all the relevant circumstances of the case. ( 135 ) It characterised that duration in some cases as ‘long’, namely ‘between 1 to more than 10 years’. ( 136 ) In that regard, it referred in particular to the GSAs with Google’s most important business partners, namely AOL and IAC. The GSAs with AOL thus covered all of its websites from mid-October 2004 to the end of 2015 – that is to say, for a duration of more than 11 years – and represented [ confidential ] % of the gross revenues generated by Google from intermediation services between 2006 and 2012. In the case of IAC, the GSAs covered all the websites from mid-May 2003 to the end of March 2016, that is, almost 13 years, and represented [ confidential ] % of the gross revenues generated by Google from those services between 2002 and 2015. This is followed by a detailed list, extending over approximately two pages, of Google’s other important direct business partners with contractual commitments having a duration of 1 to 7 years. According to the list, the GSA with Nextag had a duration of approximately 6 years and 10 months. The durations were approximately six years in the cases of British Telecom and Tripadvisor and between 4 and almost 5 years in total for Italiaonline, Kayak, ABC Startsiden and Benchmark, while CNET was committed for approximately 3 years and 10 months. The other GSAs with all sites direct partners had a shorter duration.

105. The Commission then pointed out, with reference to Section 7.3.1 ( 137 ) of the contested decision, that the fact that the exclusivity clause prevented access by competing intermediaries to a significant part of the market at issue was consistent with the evolution of Google’s market shares. ( 138 )

106. Furthermore, the Commission relied on a Microsoft study from 2011 and 2013. According to that study, the intermediation services provided on the basis of the GSAs concluded with the all sites direct partners – in terms of data traffic – concerned 93% to 100% of the most important web domains in France, Germany, Italy, Spain and the United Kingdom in 2010 and 2013. ( 139 )

107. In response to various objections raised by Google, the Commission replied that the prevention of access which had been found was in any case proved taking into account only the GSAs with the largest three all sites direct partners, AOL, eBay and IAC. ( 140 ) Between 2006 and 2009, those GSAs represented – in terms of the gross revenues generated by Google from intermediation services – between [ confidential ] % and [ confidential ] % of the total value of the market. ( 141 ) The Commission added that the gross revenues generated by Google in that regard also represented a significant percentage of the value of the market in respect of all of the all sites direct partners, namely between [ confidential ] % and [ confidential ] % from 2010 to 2012. ( 142 )

108. In response to Google’s objection reproduced in point 5 of recital 392 that the typical duration of the GSAs concerned was only two years and that certain GSAs, such as those with AOL, Orange and Tesco, had provided for unilateral termination rights, the Commission replied specifically in recitals 398 and 399 of the contested decision. Recital 398 refers in turn to recital 388, according to which the duration of the contractual commitment established by the exclusivity clause was ‘long’, namely ‘between 1 to more than 10 years’ and (certain) all sites direct partners extended many of their GSAs, sometimes frequently, without substantial modification. Footnote 570 contains three examples of GSAs each having up to three extensions. In recital 399, the Commission further maintained that a unilateral termination right did not prevent the actual application of the exclusivity clause until such time as it had been exercised. ( 143 )

(2) Placement clause

109. With regard to the placement clause, the Commission likewise assessed the duration provided for in the GSAs as part of its analysis of all the relevant circumstances of the case. ( 144 ) That assessment followed the analysis of the deterrent effect of that clause on direct partners covering a significant part of their demand from competing online advertising services intermediaries ( 145 ) and the resulting prevention of competitors accessing the market. ( 146 ) In that connection, the Commission took into consideration the gross revenues generated by Google from those services to its direct partners and all sites direct partners between 2009 and 2015, which represented [ confidential ] % to [ confidential ] % and [ confidential ] % to [ confidential ] % of the total value of the market. ( 147 ) In addition, here too, it referred to the fact that the GSAs concerned covered some of the most visited websites in the EEA (see point 106 above). ( 148 ) Furthermore, competitors were thereby prevented from accessing significant revenues. ( 149 ) The proportion of search queries generated by direct partners was also very high, as illustrated by the summary for Google’s 20 most important direct partners in the United Kingdom, Germany, France, Spain and Italy, with 73%, 89%, 79%, 65% and 70% respectively. ( 150 )

110. In recital 519 of the contested decision, the Commission found that the ‘average duration of GSAs with direct partners’ including the placement clause was ‘long’ and that many of them extended their respective GSAs, sometimes several times, without substantial modifications. Footnote 707 contains four such examples. Thus, the GSA with Tesco, after being renewed twice, was valid from 1 November 2011 to 31 March 2017, that is, for more than five years; the GSA with Trovit, after being renewed once, was valid for slightly over four years; and the GSAs with How to Media and PriceMinister, after being renewed twice (and also after being amended in the former case), were valid for almost six years.

111. Lastly, the fact that, inter alia, the placement clause helped to prevent access by competing intermediaries to the market at issue was confirmed by the evolution of shares in that market. ( 151 )

112. In recital 521 et seq. of the contested decision, the Commission rejected various arguments raised by Google. Thus, in recital 525, it responded to Google’s objection that the GSAs containing the placement clause had only a short duration and that at least one direct partner, namely Telecom Italia, had a unilateral termination right. The Commission dismissed that objection, inter alia with reference to recital 519, on the ground that the ‘average duration’ of the GSAs in question and the period during which the direct partners were required to reserve the most prominent space on their results pages for a minimum number of Google search ads were ‘long’. In that regard, footnote 713 lists a number of GSAs with direct partners which had a duration of at least three to a little over four years. In addition, the unilateral termination right in the GSA with Telecom Italia was an exception. ( 152 ) Lastly, the Commission found that Google’s stable market share and the average duration of the GSAs with direct partners containing the placement clause contradicted Google’s claim that competition was dynamic. ( 153 )

(3) Prior authorisation clause

113. The Commission also assessed the duration provided for in the GSAs, among other things, with regard to the prior authorisation clause as part of its analysis of all the relevant circumstances of the case. ( 154 ) That assessment followed its findings that the prior authorisation clause necessitated burdensome triangular negotiations between the direct partners, Google and its competing online advertising services intermediaries, that Google had refused to discuss their scope and those partners could have sourced their requirements more easily from other intermediaries without that clause. ( 155 ) In addition, the Commission found that that clause prevented access by competing intermediaries to a significant part of the market at issue. ( 156 ) As grounds, it stated inter alia that between 2011 and 2015 Google generated between approximately [ confidential ] % and [ confidential ] % of gross revenues from all direct partners ( 157 ) and between [ confidential ] % and [ confidential ] % of the total value of the market at issue ( 158 ) on the basis of the GSAs with direct partners containing that clause. Furthermore, those GSAs covered some of the most visited websites in the EEA ( 159 ) and direct partners generated high search query volumes constituting a large part of those queries in the EEA. ( 160 )

114. In recital 590, the Commission found, with reference to recital 519 of the contested decision, that the ‘average duration of the GSAs with direct partners’ containing the prior authorisation clause was ‘long’. Furthermore, the fact that the prior authorisation clause prevented access by competing intermediaries to a significant part of the market at issue was consistent with the evolution of shares in that market. ( 161 )

115. In reply to Google’s objection that the GSAs with direct partners containing the prior authorisation clause had only a short duration, ( 162 ) in recital 594 of the contested decision the Commission reiterated the finding that those GSAs had a long duration. In order to demonstrate this, it mentioned in footnote 766 a total of 11 GSAs with a duration of three years to just over four years. It also pointed out that Google and its direct partners had also extended many of those GSAs several times without substantial modifications. Footnote 767 refers in that regard to several examples where extensions were made three to four times. Lastly, the Commission found in recital 595 that Google’s stable market share and the average duration of its GSAs with direct partners contradicted Google’s claim that competition was dynamic.

(b) Examination of consideration of duration by the General Court

116. The General Court considered that assessment in the contested decision to be erroneous in respect of all the clauses at issue on grounds of an insufficient examination of the relevant circumstances of the case and failure to demonstrate exclusionary effects and annulled that decision to that extent. ( 163 )

(1) Exclusivity clause

117. With regard to the exclusivity clause, the General Court criticised the Commission, in essence, for being limited to determining the ‘total period’ or ‘cumulative duration’ (including voluntary extensions) during which the all sites direct partners were tied to Google by the GSAs and for taking into consideration insufficiently their unilateral termination rights. ( 164 ) Rather, Google had correctly pointed out inter alia the short duration of the GSAs between May 2011 and May 2013, accounting for [ confidential ] % of the total AFS turnover from direct partners in the EEA. ( 165 ) In the view of the General Court, the Commission had failed to take into consideration all the relevant circumstances of the case in the context of the assessment of the duration of the commitment under the exclusivity clause, even assuming that all of the all sites direct partners identified in the contested decision had been under such a commitment. ( 166 ) Consequently, the Commission had not demonstrated to the requisite legal standard that that clause was capable of preventing Google’s competing intermediaries from accessing a significant part of the market for online search advertising intermediation in the EEA for the duration for which it applied. ( 167 )

(2) Placement clause

118. The General Court’s assessment with regard to the duration of the GSAs in connection with the placement clause is essentially similar to that for the exclusivity clause. In that regard, the Commission had taken into consideration only the ‘total duration’, including possible extensions, or the ‘cumulative duration’, but it had taken into account ‘neither the initial duration of each of the said GSAs, taken individually, nor the duration of each of their extensions, where applicable’. ( 168 )

119. Furthermore, the Commission had relied only on the duration of GSAs concluded with at most 15 direct partners. At least 10 direct partners, which could not be regarded as an exception, enjoyed a unilateral termination right. The Commission could not therefore, ‘without having examined the actual conditions and the terms under which the extensions of the GSAs had been agreed, as well as the substance of the clauses providing for the unilateral termination rights held by some of the direct partners and the conditions in which those rights could be exercised, exclude that those … partners had the option of sourcing from Google’s competing intermediaries in respect of the most prominent spaces of their results pages, including before any extension of their GSAs, or before a unilateral termination right had been exercised’. Accordingly, the Commission also could not find that ‘those intermediaries had not had the possibility of disputing the part of the market for online search advertising intermediation covered by GSAs containing the placement clause for the total duration of those GSAs, let alone for their cumulative duration’. ( 169 )

120. According to the General Court, the Commission therefore neither took into consideration all the relevant circumstances of the case nor demonstrated to the requisite legal standard that the placement clause was capable of preventing Google’s competing intermediaries from accessing a significant part of the market for online search advertising intermediation in the EEA for the duration for which that clause applied. ( 170 )

(3) Prior authorisation clause

121. The General Court’s assessment with regard to the duration of the GSAs containing the prior authorisation clause is similar, in particular, to that in connection with the placement clause. In its analysis of average duration, the Commission had relied only on the fact that certain GSAs had been extended, sometimes several times, without substantial modifications. It had thus taken into account the total duration of the GSAs containing the prior authorisation clause by including all of their extensions, where applicable, but it had taken into account neither the initial duration of each of the said GSAs, taken individually, nor the duration of each of their extensions, where applicable. ( 171 )

122. In addition, in its analysis of the effects of the prior authorisation clause, the Commission made no mention of the unilateral termination rights held by some of the direct partners concerned. It had relied on the duration of GSAs concluded with only 11 direct partners, even though at least 5 of them had held a unilateral termination right. That was not therefore a rare exception which would justify its not being taken into account for the purposes of the assessment of the foreclosure effect of the prior authorisation clause. ( 172 )

123. In those conditions, the Commission could not, without having examined, first, the actual conditions and the terms under which GSA extensions had been agreed, second, the substance of the clauses providing for the unilateral termination rights held by some of the direct partners and, third, the conditions in which those rights could be exercised, exclude that those direct partners had the option of sourcing from Google’s competing intermediaries without having to request prior authorisation from it to change the display of competing ads, including before any extension of their GSAs or before a unilateral termination right had been exercised. Nor could it therefore find that those intermediaries had not had the possibility of disputing the part of the market for online search advertising intermediation covered by GSAs containing the prior authorisation clause for the total duration of those GSAs, let alone for their cumulative duration. ( 173 )

124. Accordingly, the Commission neither took into consideration all the relevant circumstances of the case nor demonstrated to the requisite legal standard that the prior authorisation clause was capable of preventing Google’s competing intermediaries from accessing a significant part of the market for online search advertising intermediation in the EEA for the duration for which that clause applied. ( 174 )

(c) Conclusions

125. In my view, the findings in the contested decision which are reproduced in point 103 et seq. above and the General Court found to be erroneous show that, with regard, inter alia, to the duration of the respective GSAs and the contractual commitments, the Commission carried out an overall assessment and, in some cases, relied on average values in order to determine the contribution of the clauses at issue to the exclusionary effects established by it.

126. That approach is consistent with the requirement described in points 73 and 74 above to assess foreclosure effects on the relevant market having regard to all similar contracts which are capable of influencing the conditions of competition on that market. Since it is necessary to determine whether there are actual or even merely potential foreclosure and exclusionary effects in the light of such an overall assessment, it seems fundamentally mistaken to require the Commission, as the General Court did (see points 117 to 124 above), to carry out a detailed examination and assessment of each of the individual underlying contracts and their application in order to establish whether and to what extent they contributed to those exclusionary effects. Instead, it is sufficient, in principle, to take into consideration cumulatively all the contracts which are significant to the functioning of competition on the relevant market and their respective durations and to establish that, having regard to their content and their application by the contracting parties in the period in question, they gave rise to or could have given rise to, or were capable of producing, such effects collectively (see above, point 73 in particular).

127. The necessity and appropriateness of such an overall assessment also follows from the fact that Google was the author of and partner in all the relevant GSAs containing the clauses at issue, such that any anticompetitive effects of those clauses could be attributed to it entirely. In that regard, as was explained in the contested decision, the Commission was able or even required to take into consideration as relevant circumstances of the case not only the formally binding duration of the GSAs, but also their possible (voluntary) extensions and the failure by certain (important) all sites direct partners to exercise unilateral termination rights. ( 175 ) As the Commission also submits, the finding of the existence of actual or even merely potential exclusionary effects does not depend on whether the underlying contracts had a mandatorily prescribed minimum duration during the period in which they applied to the contracting parties or were merely voluntarily extended or not terminated by those parties. In connection with the functioning of digital markets, the Court of Justice has thus recently correctly held that in assessing the existence of exclusionary effects, regard must be had above all to the economic reality. ( 176 )

128. In my view, the contested decision (notwithstanding some infelicitous phrasing ( 177 )) does satisfy those requirements governing the required individual assessment of the criterion of duration and the overall assessment of the Intel criteria, contrary to the view taken by the General Court. Indeed, it explicitly addresses – either in the main text, in particular in tables, or in the footnotes – the GSAs identified by the Commission in the administrative procedure, especially those with Google’s most important direct partners and assesses them with regard to their duration as part of a comprehensive analysis of whether the Intel criteria are satisfied. There can be no objection to the relevant findings made concerning the clauses at issue in recital 364 et seq. (exclusivity clause), ( 178 ) recital 496 et seq. (placement clause) ( 179 ) and recital 574 et seq. (prior authorisation clause) ( 180 ) of that decision, contrary to the view taken by the General Court.

129. That applies above all to the consideration of the particularly long duration of the GSAs containing the exclusivity clause with Google’s most important (including for the purposes of assessing market coverage and the functioning of competition on the relevant market) business partners. With regard to the placement clause too, the General Court criticises the Commission, incorrectly in my view, for taking into account neither the initial duration of each of the GSAs concerned nor the duration of each of their possible extensions, where applicable, ( 181 ) but merely relied on the duration of GSAs concluded with at most 15 direct partners, but not GSAs concluded with all those partners.

130. Contrary to the view taken by the General Court (see points 117 to 124 above) and by Google, it was not necessary to carry out an even closer examination of those relevant circumstances, in particular of each of the GSAs with less important business partners, which may have little or no significance or representativeness, including for the purposes of assessing market coverage and the functioning of competition on the relevant market. In that regard, as was confirmed at the hearing, moreover, neither the General Court in the judgment under appeal nor Google alleges that the Commission overlooked or disregarded certain important GSAs or contracts with competing intermediaries, ( 182 ) which would have made it possible to call into question the assessment of the resulting exclusionary effects in the contested decision. According to the (erroneous) view of the General Court, the Commission instead failed to comply with its obligation to take into consideration sufficiently the relevant circumstances of the case with regard to the exclusivity clause, even assuming that all of the all sites direct partners identified in the contested decision had been obliged to source all or most of their requirements exclusively from Google. ( 183 )

131. Both in the administrative procedure and at first instance, Google had essentially criticised only the Commission’s allegedly superficial analysis and argued that the relatively short duration of many GSAs had not prevented the direct partners bound by them from approaching competing intermediaries for their entire (contestable) demand. At the hearing too, Google reiterated, in essence, that those intermediaries had the opportunity, after the GSAs had expired or been terminated, to compete for the demand of the direct partners, which should have led the Commission to reject the existence of exclusionary effects arising from the clauses at issue. Because of the short duration of the contractual commitment in certain cases, this had at times affected almost 90% of the volume of demand. Google had always acknowledged, however, that, even though that demand was contestable, the competing intermediaries had remained largely unsuccessful with their services, which were allegedly inferior to those offered by Google. Nor was Google able to demonstrate whether and to what extent the expiry of the GSAs, which, according to it, came up for renewal between May 2011 and May 2013 and accounted for [ confidential ] % of the total AFS turnover from direct partners in the EEA, ( 184 ) had actually resulted in a change of provider. ( 185 )

132. The existence of an error of law by the General Court becomes all the more apparent if the assessment of the duration of the GSAs contained in the contested decision is considered in conjunction with the assessments of the other relevant ( Intel ) criteria for establishing a foreclosure or exclusionary effect in the period in question, which the General Court largely confirmed as being free from errors of assessment and errors of law and which, in the absence of a cross-appeal brought by Google, have become final. That applies, in particular, to the assessment of, first , the evolution of Google’s market shares, second , the deterrent effect of the clauses at issue on direct partners approaching competing intermediaries, third , the market coverage by the GSAs and, fourth , the resulting prevention of those intermediaries from accessing a significant part of the market at issue. ( 186 ) Furthermore, as the General Court also recognised in principle, except as regards duration, those circumstances and effects are mutually reinforcing, including in the interrelationship between the clauses at issue. The assessment in the contested decision in that regard and the confirmatory findings in the judgment under appeal in fact indicate that, in the view of the General Court too, the Commission had succeeded, at least partially, in demonstrating even actual (that is to say, not merely potential) exclusionary effects produced by the clauses at issue.

133. Particularly relevant factors in that respect are the evolution of Google’s market shares to the detriment of its competitors while the GSAs were valid and applicable in the period in question ( 187 ) and the fact that the relevant markets were characterised by significant barriers to entry and expansion, scale and network effects and the lack of competitors and countervailing buyer power. As was recognised by the General Court – and is not contested by Google – the Commission was entitled, against that background, to conclude that exclusionary effects existed in view of the deterrent effect of the clauses at issue on its direct partners approaching competing intermediaries, the market coverage by the GSAs ( 188 ) and the resulting prevention of those intermediaries from accessing a significant part of the market at issue. That applies to all the clauses at issue to the end of 2015 at least. The General Court, in essence, definitively endorsed that assessment, except as regards duration, ( 189 ) in the judgment under appeal. ( 190 ) Having regard to the required overall assessment of those circumstances, the General Court’s criticism that the Commission had not, on account of its allegedly insufficient examination of the criterion of duration, demonstrated to the requisite legal standard that collectively the clauses at issue produced or were at least capable of producing exclusionary effects also therefore appears to be contradictory. ( 191 )

134. In any event, the General Court could not, without committing an error of law, criticise the Commission for failing to examine sufficiently or to assess and weigh – in their overall context – the various relevant circumstances and criteria for assessing possible exclusionary effects. Nor can it be inferred from the General Court’s assessment, reproduced above, ( 192 ) of whether and to what extent the Commission’s findings on Google’s market position, the structure and functioning of the relevant markets and the market coverage by the clauses at issue from 2006 to 2015 and their capability to foreclose competing intermediaries, which were confirmed by the General Court (and have thus become final), might have been qualified by the allegedly insufficient examination and assessment of the duration of the GSAs and of those clauses. Conversely, the General Court does not explain in the judgment under appeal whether and to what extent a more detailed analysis of that duration could have altered the outcome of that assessment, positively or negatively, all the more so as the (cumulative) duration and/or period of application of the clauses at issue, in particular the exclusivity clause, but also the placement clause, with many of Google’s important business partners was considerable.

135. Those conclusions cannot be invalidated by the central argument repeatedly advanced by Google that, on account of the short contractual commitments, the volume of demand tied to it was contestable by the competing intermediaries, whose efforts had failed, however, due to the inferior quality of their offering. In so far as Google thus claims that the exclusionary effects found by the Commission cannot be attributed to the clauses at issue, that disregards the many indications set out above on which the Commission – now uncontested by Google – relied (and was entitled to rely) in order to arrive at the opposite assumption. In that respect, the question also raised by the Commission in these proceedings does indeed arise: why was it actually necessary, from Google’s perspective, despite the allegedly superior quality of its offering, to introduce those clauses and to continue to apply them on a graduated basis for years? In view of the overall context established in the contested decision and definitively confirmed in the judgment under appeal, Google cannot in any event rely on that argument to call into question the causal link and the attribution link, accepted by the Commission and acknowledged by the General Court, between the clauses at issue, on the one hand, and the – at least potential, if not actual – exclusionary effects and restrictions on competition, on the other hand.

136. In my view, that error of law on the part of the General Court, which can be established in the context of the assessment of the second, third and fourth pleas in law throughout the judgment under appeal, means that the first, second and third grounds of appeal must each be upheld in respect of most of the complaints raised by the Commission. The following remarks therefore address those complaints only briefly and in so far as they have not yet been answered by the preceding analysis.

D. Grounds of appeal in detail

1. First ground of appeal regarding the exclusivity clause

137. In my view, the three branches of the first ground of appeal, which overlap considerably in substance, are largely well founded for the reasons set out above.

(a) First branch: complete assessment of the Intel criteria and of the relevant circumstances of the case

138. It follows from the statements made in point 71 et seq. above that the General Court incorrectly criticised the Commission, with regard to the assessment whether the Intel criteria, in particular the criterion of duration, were satisfied, for carrying out an incomplete examination of the restrictive effects on competition of the exclusivity clause (see point 116 et seq. above).

139. Rather, the Commission had, as required by the case-law of the Court of Justice, ( 193 ) analysed, weighed and interrelated those criteria sufficiently, taking into consideration all the relevant circumstances of the case in order to establish that in the period from 2006 to 2015 that clause produced the exclusionary effects found in the contested decision.

(b) Second branch: sufficient demonstration of exclusionary effects

140. It also follows that, as was explained in point 116 et seq. above, the General Court relied on an incorrect legal standard in assessing the relevant circumstances of the case and in demonstrating them (standard of proof) in order to establish the existence of exclusionary effects in the period from 2006 to 2015. As the Commission has submitted, the General Court further failed to explain in the judgment under appeal whether and to what extent the more detailed examination, required by it, of the duration of the GSAs containing the exclusivity clause could have called into question the very exclusionary effects which it had itself confirmed, as they had been established in the contested decision. In my view, however, that does not necessarily lead to a finding of a (formal) failure to state reasons, since it is perfectly possible for the Court of Justice, on the basis of the statement of reasons in the judgment under appeal, to review and establish in that regard the existence of an error of law on the part of the General Court (that is, in relation to the merits). ( 194 )

(c) Third branch: inter alia undue reversal of the burden of proof to the detriment of the Commission with regard to the assessment of the duration of the exclusivity clause

141. The many errors of law which the General Court is alleged to have committed in the context of third branch of the first ground of appeal ( 195 ) do not require a detailed or separate examination, in my view, but largely overlap with the complaints already addressed in connection with the first and second branches of the first ground of appeal. That concerns in particular the Commission’s – alleged – obligation to carry out a detailed assessment with regard to the (only short) duration, the termination right and the extensions of each GSA, and the distortion of the content of the contested decision and the Commission’s arguments at first instance claimed in that respect, the General Court’s contradictory or substituting reasoning and the undue reversal of the burden of proof to its detriment, as well as the misapplication of Article 102 TFEU in the light of paragraph 36 of the Priorities Communication.

142. In particular, the complaint of the undue reversal of the burden of proof is likewise well founded in view of the incorrect standard of review and standard of proof applied by the General Court and the fact that, according to the other findings in the judgment under appeal, the Commission had adequately established not only potential, but even actual exclusionary effects of the clauses at issue. That also applies to the complaint of contradictory reasoning on the part of the General Court since, despite the exclusionary effects of the exclusivity clause largely confirmed by it, it called into question the finding of such effects solely on the basis of the allegedly insufficient examination of the criterion of duration in order to annul the contested decision to that extent.

143. In other respects, however, the complaints of distortion and failure to give a proper statement of reasons are, in my view, largely ineffective. In that regard, the General Court did not so much reproduce the content of the contested decision or of the Commission’s submissions incorrectly as applied a legally flawed premiss in relation to the scope of the obligation to take into consideration and examine the relevant circumstances of the case. It was primarily on the basis of that false premiss that it considered the Commission’s examination of those circumstances and the associated reasoning in the contested decision regarding the criterion of duration – including in cumulative form – to be insufficient, even though that decision contained all the necessary assessments and evidence in that regard. In any event, it is sufficient to find such an error of law on the part of the General Court, without it being necessary, in turn, to characterise the classification and examination of the relevant facts and evidence, which are necessarily also flawed as a result, as a distortion or as an error of law on other grounds. I therefore consider that, if the Court of Justice concurs with my proposals, it should refrain from responding to those rather piecemeal complaints ( 196 ) raised by the Commission.

144. That also makes it unnecessary to address Google’s repeatedly advanced but largely incorrect objection that those complaints are aimed, in part, at obtaining a fresh assessment of facts or evidence, some of which were submitted out of time by the Commission, or a ‘rewriting’ of the contested decision by the EU Courts, ( 197 ) which, in principle, falls outside of the jurisdiction of the Court of Justice for review in appeal proceedings. ( 198 )

145. Consequently, all three branches of the first ground of appeal must be upheld.

2. Second ground of appeal regarding the placement clause

146. For the abovementioned reasons, the three branches of the second ground of appeal are likewise largely well founded.

(a) First branch: complete assessment of the Intel criteria and of the circumstances of the case

147. The succinctly formulated first branch of the second ground of appeal refers, in essence, to the first branch of the first ground of appeal with the result that it should likewise be upheld on the basis of the considerations set out in points 138 and 139 above concerning the assessment of the exclusionary effects of the placement clause. ( 199 )

(b) Second branch: sufficient demonstration of exclusionary effects

148. In support of the second branch of the second ground of appeal, the Commission claims, in essence, that the General Court had largely – except for duration – confirmed definitively the assessment of the Intel criteria in the contested decision and the establishment of the resulting exclusionary effects. However, it erred in law in failing to explain why the allegedly inadequate examination of duration alone could call into question the outcome of that overall analysis of the relevant circumstances of the case regarding the capability of the placement clause to restrict competition.

149. I can only agree with those arguments in the light of my remarks concerning the exclusionary effects of the placement clause (see point 125 et seq. above), as they are described in the contested decision. ( 200 ) In that regard too, it is no longer relevant whether and to what extent that conclusion might be precluded by the second Intel judgment of the Court of Justice ( 201 ) (see point 144 above).

(c) Third branch: inter alia undue reversal of the burden of proof to the detriment of the Commission with regard to the assessment of the duration of the placement clause

150. With respect to the third branch of the second ground of appeal, I refer to my above comments in points 141 to 145. In that regard, the Commission likewise advances a number of complaints which overlap with those in the first and second branches of the second ground of appeal and are very similar to those in the third branch of the first ground of appeal. ( 202 ) Accordingly, the claim of the reversal of the burden of proof to the detriment of the Commission is also well founded in this connection. However, for the reasons already set out above, it is not necessary to respond separately to each individual complaint (concerning distortion of the content of the contested decision or the Commission’s submissions).

3. Third ground of appeal regarding the prior authorisation clause

151. The assessment of the third ground of appeal should be exactly the same as for the first and second grounds of appeal. It should also therefore be upheld.

(a) First branch: complete assessment of the Intel criteria and of the circumstances of the case

152. With regard to the prior authorisation clause, the Commission’s arguments in relation to the first branch of the third ground of appeal are well founded in accordance with the statements made in points 138 and 139 above. The General Court erred in law in considering the findings and evidence set out in the contested decision concerning the existence of (at least potential) exclusionary effects arising from that clause ( 203 ) to be insufficient.

(b) Second branch: sufficient demonstration of exclusionary effects

153. It follows that the second branch of the third ground of appeal concerning the examination of the relevant circumstances of the case and demonstration of the satisfaction of the Intel criteria in the contested decision in respect of the prior authorisation clause ( 204 ) is also well founded.

(c) Third branch: inter alia undue reversal of the burden of proof to the detriment of the Commission with regard to the assessment of the duration of the prior authorisation clause

154. The third branch of the third ground of appeal contains a number of complaints which overlap with those in the first and in the second branches of that ground of appeal and are very similar to those in the third branches of the first and second grounds of appeal. ( 205 ) For the reasons set out in points 141 to 145 above, the claim of the reversal of the burden of proof to the detriment of the Commission is likewise well founded in this connection. However, it is not necessary to respond separately to each individual complaint (concerning distortion of the content of the contested decision or the Commission’s submissions).

4. Interim conclusion

155. The first, the second and the third grounds of appeal must therefore each be upheld.

5. Fourth ground of appeal: single and continuous infringement

156. It follows from the statements regarding the first, the second and the third grounds of appeal that the General Court erred in law by criticising the Commission for failing to demonstrate the existence of exclusionary effects arising, inter alia, from each of the clauses at issue. Accordingly, its finding that the infringements based on each of those clauses could not constitute a single and continuous infringement – as established in the contested decision ( 206 ) – also necessarily constitutes an error of law. ( 207 )

157. In that regard, I would draw attention, additionally, to the clarification recently provided in the case-law of the Court of Justice according to which the objective and subjective criteria for determining the existence of a single and continuous infringement also apply in the context of the application of Article 102 TFEU, ( 208 ) in particular where the undertaking in a dominant position thereby pursues an overall strategy to foreclose competitors. ( 209 ) As the Commission reaffirmed at the hearing, it is plausible that such a strategy was pursued here, but that would have to be assessed more closely by the General Court after the case has been referred back to it, as I recommend (see point 161 et seq. below). That is the only explanation for the successive shift, initiated by Google, in the design of the GSAs, from the exclusivity clause to the placement and prior authorisation clauses, in the period from 2006 to 2015. In any event, the ostensible economic rationale behind, and need for, those clauses are difficult to reconcile with the superior quality of the services which Google claims to have offered and a practice guided solely by competition on the merits (see point 135 above).

158. The fourth ground of appeal must therefore also be upheld.

E. Referral back to the General Court

159. Under Article 61 of the Statute of the Court of Justice of the European Union, the Court of Justice must quash the decision of the General Court if the appeal is well founded. It may itself give final judgment in the matter, where the state of the proceedings so permits, or refer the case back to the General Court for a fresh judgment.

160. It follows from all the foregoing considerations that the appeal is well founded and that the judgment under appeal must be set aside in so far as it annulled the contested decision on grounds of failure to demonstrate exclusionary effects in the period from 2006 to the end of 2015.

161. Aside from the individual complaints in connection with the second, third and fourth pleas in law which were not addressed, however, in the judgment under appeal the General Court examined neither the admissibility and merits, which were contested by the Commission, of Google’s claim challenging the classification of its alleged practice as a single and continuous infringement, nor the fifth plea in law, which was directed against the imposition of the fine.

162. With regard to the imposition of the fine by the Commission in particular, in accordance with Article 261 TFEU and Article 31 of Regulation No 1/2003, the General Court has unlimited jurisdiction. It is thus empowered, in addition to carrying out a mere review of the lawfulness of those fines, to substitute its own appraisal for that of the Commission and, consequently, to cancel, reduce or increase the fine imposed, where that has been applied for. ( 210 ) On the other hand, it is not for the Court of Justice, in an appeal, ‘to substitute, on grounds of fairness, its own assessment for that of the General Court’, but may find the existence of an error of law only inasmuch as ‘the level of the penalty [imposed by the General Court] is not merely inappropriate, but also excessive to the point of being disproportionate’. ( 211 )

163. Having regard to the latter requirement in particular, I do not consider that the state of the proceedings permits judgment to be given since, if it were to give judgment itself, the Court of Justice would have to deal, for the first time, with Google’s detailed complaints regarding the method applied by the Commission in calculating the fine and the proportionality of that fine, which fall primarily within the jurisdiction of the General Court as the court responsible for the assessment of facts. ( 212 ) Moreover, the fine imposed in the contested decision must be reduced in any event, as the Commission has acknowledged the failure to demonstrate the existence of exclusionary effects in respect of the clauses at issue in 2016 or has not challenged the General Court’s assessment in that regard.

164. On those grounds, the Court of Justice cannot, in my view, give final judgment on all the pleas in law and complaints raised at first instance, and on the fifth plea in law in particular.

165. The case must therefore be referred back to the General Court, the costs being reserved.

IV. Conclusion

166. I therefore propose that the Court should:

– Set aside the judgment of 18 September 2024, Google and Alphabet v Commission (Google AdSense for Search) (T‑334/19, EU:T:2024:634);

– Refer the case back to the General Court of the European Union;

– Reserve the costs.

1 Original language: German.

2 Judgment of 18 September 2024, Google and Alphabet v Commission (Google AdSense for Search) (T‑334/19, EU:T:2024:634).

3 Decision C(2019) 2173 final of 20 March 2019 relating to a proceeding under Article 102 [TFEU] and Article 54 of the EEA Agreement (Case AT.40411 – Google Search (AdSense)).

4 See, in particular, judgment of 21 December 2023, European Superleague Company (C‑333/21, EU:C:2023:1011, paragraphs 119 and 186).

5 Judgment of 6 September 2017 (C‑413/14 P, EU:C:2017:632).

6 The judgment in Intel I (paragraph 139).

7 See paragraphs 2 to 15 of the judgment under appeal.

8 For that reason, reference is also made to two-sided (digital) markets, which have been familiar for some time, in particular through advertising-supported television; see the overview in OECD Competition Committee, Two-Sided Markets , Series Roundtables on Competition Policy No 97, 2009 (available at: Two-Sided Markets | OECD); see also Opinion of Advocate General Collins in Booking.com and Booking.com (Deutschland) (C‑264/23, EU:C:2024:470, point 69).

9 See recitals 2, 13 and 25 of Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on contestable and fair markets in the digital sector and amending Directives (EU) 2019/1937 and (EU) 2020/1828 (Digital Markets Act) (OJ 2022 L 265, p. 1): ‘extreme scale economies, … very strong network effects, an ability to connect many business users with many end users through the multisidedness of these services, a significant degree of dependence of both business users and end users, lock-in effects, a lack of multi-homing for the same purpose by end users, vertical integration, and data driven-advantages’.

10 See recitals 16 to 20 of the contested decision as well as judgment of 5 August 2024 of the United States District Court for the District of Columbia, Case No. 20-cv-3010 (APM), paragraph 172 et seq. and paragraph 238 et seq. (available at: Google Search Engine Monopoly Ruling.pdf).

11 An overview of the various intermediation services can be found in recital 21 et seq. of the contested decision.

12 Commission Regulation of 7 April 2004 relating to the conduct of proceedings by the Commission pursuant to Articles [101] and [102 TFEU] (OJ 2004 L 123, p. 18).

13 Council Regulation of 16 December 2002 on the implementation of the rules on competition laid down in Articles [101] and [102 TFEU] (OJ 2003 L 1, p. 1).

14 See recital 222 et seq. and recital 274 et seq. of the contested decision; paragraph 39 et seq. of the judgment under appeal.

15 Recital 340 et seq. of the contested decision; paragraphs 67 to 73 of the judgment under appeal.

16 Judgment of 13 February 1979 (85/76, EU:C:1979:36).

17 Recitals 340 to 361 of the contested decision. See the judgment in Hoffmann-La Roche (paragraph 89), according to which ‘an undertaking which is in a dominant position on a market and ties purchasers – even if it does so at their request – by an obligation or promise on their part to obtain all or most of their requirements exclusively from the said undertaking abuses its dominant position within the meaning of Article [102 TFEU]’.

18 Recital 362 et seq. of the contested decision; summarised in paragraphs 71 to 73 of the judgment under appeal.

19 Recital 437 et seq. of the contested decision; summarised in paragraphs 74 to 77 of the judgment under appeal.

20 Recital 459 et seq. of the contested decision; summarised in paragraph 78 et seq. of the judgment under appeal.

21 Recitals 553 to 562 of the contested decision; summarised in paragraphs 83 to 86 of the judgment under appeal.

22 Recital 566 et seq. of the contested decision; summarised in paragraph 87 et seq. of the judgment under appeal.

23 Recital 615 et seq. of the contested decision; summarised in paragraphs 90 to 92 of the judgment under appeal.

24 Recital 625 et seq. of the contested decision; summarised in paragraphs 93 to 95 of the judgment under appeal.

25 Paragraphs 114 to 369 of the judgment under appeal.

26 Paragraphs 1003 to 1005 of the judgment under appeal.

27 Judgment of 19 January 2023 (C‑680/20, EU:C:2023:33).

28 Paragraphs 379 to 391 of the judgment under appeal.

29 Paragraphs 398 to 727 of the judgment under appeal.

30 Paragraphs 409 to 682 of the judgment under appeal.

31 Paragraphs 717 and 722 to 724 of the judgment under appeal.

32 Paragraphs 724 to 727 of the judgment under appeal.

33 Paragraphs 732 to 768 of the judgment under appeal.

34 Paragraph 748 of the judgment under appeal.

35 Paragraphs 767 and 768 of the judgment under appeal.

36 Paragraphs 769 to 913 of the judgment under appeal.

37 Paragraph 907 of the judgment under appeal.

38 Paragraphs 749 to 883 and 902 of the judgment under appeal.

39 See also paragraphs 908 and 909 of the judgment under appeal.

40 Paragraphs 910 and 911 of the judgment under appeal.

41 Paragraphs 912 and 913 of the judgment under appeal.

42 See paragraph 923 of the judgment under appeal.

43 Paragraphs 914 to 1002 of the judgment under appeal.

44 Paragraphs 929 to 981 and 992 of the judgment under appeal.

45 Paragraphs 997 to 999 (with reference to paragraph 993) of the judgment under appeal. The latter aspect does not form part of the appeal.

46 Paragraphs 1000 to 1002 of the judgment under appeal.

47 Order of the President of the Court of 22 May 2025, Commission v Google and Alphabet (C‑826/24 P, EU:C:2025:386).

48 Order of 30 June 2025, Commission v Google and Alphabet (C‑826/24 P, EU:C:2025:538).

49 The approach taken by the General Court in that regard recalls the Opinion of Advocate General Wahl in Intel v Commission (C‑413/14 P, EU:C:2016:788, points 147 to 157).

50 The judgment in Intel I (paragraph 139).

51 See, to that effect, the judgment in Intel I (paragraphs 137 to 139), and judgment of 24 October 2024, Commission v Intel Corporation (C‑240/22 P, EU:C:2024:915, paragraphs 130, 144, 180, 273 and 331). See also judgments of 12 May 2022, Servizio Elettrico Nazionale and Others (C‑377/20, EU:C:2022:379, paragraph 51), and of 10 September 2024, Google and Alphabet v Commission (Google Shopping) (C‑48/22 P, EU:C:2024:726, paragraph 265).

52 See, on the corresponding obligations with regard to the demonstration incumbent on Member States’ competition authorities, judgment of 19 January 2023, Unilever Italia Mkt. Operations (C‑680/20, EU:C:2023:33, paragraph 48 et seq.).

53 Paragraphs 379 and 380 of the judgment under appeal.

54 See, to that effect, the judgment in Hoffmann-La Roche (paragraph 89).

55 See, to that effect, the judgment in Intel I (paragraphs 111 and 129 et seq., with reference to the circumstances relevant to a loyalty rebate scheme, such as the level of the rebates in question, their duration, the market shares concerned, the needs of customers and the capability of the rebates to foreclose an as-efficient competitor); and judgments of 12 May 2022, Servizio Elettrico Nazionale and Others (C‑377/20, EU:C:2022:379, paragraph 72), and of 19 January 2023, Unilever Italia Mkt. Operations (C‑680/20, EU:C:2023:33, paragraphs 40, 41, 44 and 62, with regard to the obligations with regard to examination and demonstration incumbent on Member States’ competition authorities).

56 See paragraph 107 et seq., and paragraphs 238, 381, 385, 388 to 390, 562, 595, 597, 614, 694 to 699, 715 to 722, 835, 843, 848, 891 to 898, 903 to 908 and 988 to 998 of the judgment under appeal.

57 Recitals 340 to 361, and recital 362 et seq. of the contested decision.

58 Recital 459 et seq. and recital 566 et seq. of the contested decision.

59 See paragraphs 698, 699, 702, 715, 721, 776, 849, 884, 893, 894, 897, 907, 926, 982, 988 to 990 and 997 of the judgment under appeal.

60 See recitals 364, 373, 388, 392, 398 and 399 (with regard to the exclusivity clause), 496, 519, 525 and 527 (with regard to the placement clause) and 574, 590, 592, 594 and 595 (with regard to the prior authorisation clause) of the contested decision.

61 With regard to the concept of ‘potential’ restrictive effects on competition, see, most recently, judgment of 5 December 2024, Tallinna Kaubamaja Grupp and KIA Auto (C‑606/23, EU:C:2024:1004, paragraphs 32 and 33 and the case-law cited).

62 See, with regard to forms of conduct which are harmful to competition ‘by their very nature’, the effects of which do not need to be established, judgments of 12 January 2023, Lietuvos geležinkeliai v Commission (C‑42/21 P, EU:C:2023:12, paragraph 76 et seq.), and of 21 December 2023, European Superleague Company (C‑333/21, EU:C:2023:1011, paragraphs 147 to 152 and 178). With regard to ‘naked restrictions’, see judgment of 26 January 2022, Intel Corporation v Commission (T‑286/09 RENV, EU:T:2022:19, paragraphs 88 to 96), and – dismissing the appeal brought by the Commission – judgment of 24 October 2024, Commission v Intel Corporation (C‑240/22 P, EU:C:2024:915, paragraphs 38, 83 and 97), and also judgment of 10 December 2025, Intel Corporation v Commission (T‑1129/23, EU:T:2025:1091) (appeal and cross-appeal pending in Case C‑143/26 P).

63 See judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping) (C‑48/22 P, EU:C:2024:726, paragraphs 165 to 168 and 265). See also judgments of 2 July 2026, Google and Alphabet v Commission (C‑738/22 P, EU:C:2026:533, paragraphs 187 to 190 and 274), and of 21 December 2023, European Superleague Company (C‑333/21, EU:C:2023:1011, paragraphs 129 to 131).

64 See, to that effect, judgment of 2 July 2026, Google and Alphabet v Commission (C‑738/22 P, EU:C:2026:533, paragraph 189 and the case-law cited; see also paragraph 191, according to which ‘the taking into account of a factual element as a relevant contextual factor for the purpose of assessing the effects of conduct in the light of Article 102 TFEU does not depend on whether that element amounts to [other] behaviour that in itself is characterised as abusive’).

65 See, to that effect, judgments of 30 January 2020, Generics (UK) and Others (C‑307/18, EU:C:2020:52, paragraph 146), and, in particular, of 21 December 2023, European Superleague Company (C‑333/21, EU:C:2023:1011, paragraphs 119 and 186).

66 See the situation underlying the judgment of 19 January 2023, Unilever Italia Mkt. Operations (C‑680/20, EU:C:2023:33), which is comparable in that respect.

67 Judgment of 28 February 1991 (C‑234/89, EU:C:1991:91, paragraphs 19 to 27). See also judgment of 26 November 2015, Maxima Latvija (C‑345/14, EU:C:2015:784, paragraphs 25 to 30, clarifying that establishing a potentially restrictive effect on competition is also sufficient).

68 See, to that effect, judgment of 28 February 1991, Delimitis (C‑234/89, EU:C:1991:91, paragraph 26).

69 Communication from the Commission – Commission notice – Guidelines on vertical restraints (OJ 2022 C 248, p. 1), paragraphs 281, 302 (‘The higher the proportion of its market share that a supplier sells under a single branding obligation and the longer the duration of the single branding obligations, the more significant foreclosure is likely to be’) and 324 (‘As well as the market position of the buyer on the upstream and downstream market, it is also important to take into account the extent and duration of the exclusive supply obligation. The higher the tied supply share, and the longer the duration of the exclusive supply obligation, the more significant the foreclosure effect is likely to be’).

70 See paragraphs 13 to 15 of the Communication from the Commission ‘Guidance on the Commission’s enforcement priorities in applying Article [102 TFEU] to abusive exclusionary conduct by dominant undertakings’ (OJ 2009 C 45, p. 7). The Commission’s understanding in that regard is also expressed in paragraph 106 of its Communication of 3 September 2026 (C(2026) 6118 final) – ‘Guidelines on the application of Article 102 [TFEU] to abusive exclusionary conduct by dominant undertakings’ (available at: Commission adopts EU Guidelines on exclusionary abuses of dominance).

71 Communication from the Commission ‘Guidance on the Commission’s enforcement priorities in applying Article [102 TFEU] to abusive exclusionary conduct by dominant undertakings’ (OJ 2009 C 45, p. 7).

72 Paragraphs 696 and 712 of the judgment under appeal.

73 See also Opinion of Advocate General Wahl in Intel v Commission (C‑413/14 P, EU:C:2016:788, point 150, according to which the short duration of an arrangement does not exclude it from having anticompetitive effects); and Communication from the Commission of 3 September 2026 (footnote 70), paragraph 106(d): ‘In general, the higher the share of total sales in the relevant market affected by the conduct and the longer the duration of the conduct, the greater is the capability of the conduct to produce exclusionary effects ’ (emphasis added).

74 See paragraph 712 of the judgment under appeal.

75 See, to that effect, judgment of 21 January 2016, Galp Energía España and Others v Commission (C‑603/13 P, EU:C:2016:38, paragraph 90). With regard to the lack of binding effect on Member States, see judgments of 13 December 2012, Expedia (C‑226/11, EU:C:2012:795, paragraph 29), and of 20 January 2016, DHL Express (Italy) and DHL Global Forwarding (Italy) (C‑428/14, EU:C:2016:27, paragraph 33).

76 A similar understanding is adopted in the Communication from the Commission of 3 September 2026 (footnote 70), paragraphs 106, 142 and 161.

77 See Article 2(2) and (3) of Council Regulation (EC) No 139/2004 of 20 January 2004 on the control of concentrations between undertakings (the EC Merger Regulation) (OJ 2004 L 24, p. 1).

78 Judgment of 13 July 2023, Commission v CK Telecoms UK Investments (C‑376/20 P, EU:C:2023:561, paragraph 262).

79 See, to that effect, judgment of 13 July 2023, Commission v CK Telecoms UK Investments (C‑376/20 P, EU:C:2023:561, paragraph 270).

80 See also, to that effect, for exclusivity commitments, judgment of 19 January 2023, Unilever Italia Mkt. Operations (C‑680/20, EU:C:2023:33, paragraphs 39 and 41 and the case-law cited). This is also referred to in paragraphs 106 and 108 of the judgment under appeal.

81 Known in English ‘as-efficient competitor test’ (AEC test). With regard to its (mandatory) application in the case of price competition, see judgment of 24 October 2024, Commission v Intel Corporation (C‑240/22 P, EU:C:2024:915, paragraphs 181, 201 and 343).

82 See, to that effect, judgment of 19 January 2023, Unilever Italia Mkt. Operations (C‑680/20, EU:C:2023:33, paragraph 57 and the case-law cited). See also paragraph 661 of the judgment under appeal. I have repeatedly made clear that such a restriction is necessary for digital markets in particular; see, for example, my Opinion in Google and Alphabet v Commission (C‑738/22 P, EU:C:2025:463, point 125 et seq.).

83 See my Opinion in Google and Alphabet v Commission (C‑48/22 P, EU:C:2024:14, points 194 to 197).

84 See, to that effect, judgment of 19 January 2023, Unilever Italia Mkt. Operations (C‑680/20, EU:C:2023:33, paragraphs 59 to 61).

85 See, to that effect, the somewhat convoluted phrasing of the Grand Chamber of the Court of Justice in the judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping) (C‑48/22 P, EU:C:2024:726, paragraph 165 and the case-law cited).

86 Judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping) (C‑48/22 P, EU:C:2024:726, paragraphs 170 to 172). I have already referred to the need for such a distinction and separate examination in my Opinion in Google and Alphabet v Commission (C‑48/22 P, EU:C:2024:14, point 144). This seems to have been recently acknowledged by the Court of Justice, which characterises them as conditions to be satisfied cumulatively ; see judgment of 2 July 2026, Google and Alphabet v Commission (C‑738/22 P, EU:C:2026:533, paragraph 264 and the case-law cited; emphasis added).

87 See, to that effect, judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping) (C‑48/22 P, EU:C:2024:726, paragraph 167 and the case-law cited; emphasis added). See also, in that regard, my Opinion in Google and Alphabet v Commission (C‑48/22 P, EU:C:2024:14, point 192 et seq.). This approach is confirmed by the judgment of 2 July 2026, Google and Alphabet v Commission (C‑738/22 P, EU:C:2026:533, paragraphs 271 to 274).

88 With regard to ‘tying’ on digital markets, which represents, in principle, means that deviate from competition on the merits, and the anticompetitive effects of which should not be assessed using the as-efficient competitor test, see judgment of 2 July 2026, Google and Alphabet v Commission (C‑738/22 P, EU:C:2026:533, paragraph 269 et seq.).

89 See, to that effect, judgments of 19 January 2023, Unilever Italia Mkt. Operations (C‑680/20, EU:C:2023:33, paragraph 38 and the case-law cited), and of 10 September 2024, Google and Alphabet v Commission (Google Shopping) (C‑48/22 P, EU:C:2024:726, paragraph 163 and the case-law cited).

90 See, to that effect, judgment of 10 September 2024, Google and Alphabet v Commission (Google Shopping) (C‑48/22 P, EU:C:2024:726, paragraph 88 and the case-law cited). See also judgment of 2 July 2026, Google and Alphabet v Commission (C‑738/22 P, EU:C:2026:533, paragraph 264 and the case-law cited, according to which a practice may be capable of producing an exclusionary effect, in that it is capable of making it more difficult for competitors to enter or remain on the market in question, and, by so doing, is capable of having an impact on the market structure ; emphasis added). See also my Opinion in Google and Alphabet v Commission (C‑738/22 P, EU:C:2025:463, points 106 and 131 to 136). The judgment of 16 March 2023, Towercast (C‑449/21, EU:C:2023:207, paragraph 52 and the case-law cited, according to which establishing the existence of abuse requires inter alia ‘that the degree of dominance thus reached would substantially impede competition, that is to say, that only undertakings whose behaviour depends on the dominant undertaking would remain in the market’), points in a similar direction (protection of market structure).

91 See, to that effect – with regard to situations in which it is not possible, nor does it make sense, to base the analysis of exclusionary effects of particular conduct on the question whether a (hypothetical) as-efficient competitor could replicate that conduct – judgment of 2 July 2026, Google and Alphabet v Commission (C‑738/22 P, EU:C:2026:533, paragraphs 273 and 278). See also, in that regard, my Opinion in Google and Alphabet v Commission (C‑738/22 P, EU:C:2025:463, point 125 et seq.).

92 Recitals 230, 276 and 364 of the contested decision.

93 See paragraphs 558 to 560, 666, 846, 882, 944 and 980 of the judgment under appeal.

94 Paragraph 562 of the judgment under appeal, with reference to recital 364 of the contested decision.

95 Paragraph 848 of the judgment under appeal, with reference to recital 496 of the contested decision.

96 See, in particular, paragraphs 944 and 978 to 981 of the judgment under appeal.

97 Recital 228 et seq. and recital 274 et seq.

98 See, in particular, recital 253 of the contested decision.

99 See, in particular, recital 280 et seq. of the contested decision.

100 Paragraph 563 et seq. (with regard to the exclusivity clause), paragraph 849 et seq. (with regard to the placement clause) and paragraph 954 et seq. (with regard to the prior authorisation clause) of the judgment under appeal.

101 Paragraph 409 et seq. (with regard to the exclusivity clause), paragraph 781 et seq. (with regard to the placement clause) and paragraph 929 et seq. (with regard to the prior authorisation clause) of the judgment under appeal.

102 According to the General Court, Google had not provided during the administrative procedure or in the proceedings before it any analysis based on the as-efficient competitor test within the meaning of the judgment in Intel I (paragraph 138 et seq.); see paragraphs 671, 881 and 979 of the judgment under appeal.

103 Paragraph 651 et seq. (with regard to the exclusivity clause), paragraph 876 et seq. (with regard to the placement clause) and paragraph 974 et seq. (with regard to the prior authorisation clause) of the judgment under appeal.

104 Paragraphs 665, 678 to 682, 716 (with regard to the exclusivity clause), 880 to 883 (with regard to the placement clause), 963 and 978 to 981 (with regard to the prior authorisation clause) of the judgment under appeal.

105 Paragraphs 409 to 562 of the judgment under appeal.

106 Paragraphs 563 to 682 of the judgment under appeal.

107 Paragraphs 571, 578, 584, 590, 592, 605, 618, 633 and 642 of the judgment under appeal.

108 Paragraphs 665, 682 and 716 of the judgment under appeal.

109 Paragraphs 558 to 560 of the judgment under appeal, with reference to the statements made in recitals 230, 276 and 364 of the contested decision.

110 Paragraph 572 et seq. of the judgment under appeal.

111 Paragraphs (562 and) 650 of the judgment under appeal, which do not form part of the appeal.

112 Paragraph 843 of the judgment under appeal.

113 Paragraph 767 of the judgment under appeal.

114 Paragraphs 839 and 840 of the judgment under appeal.

115 Paragraphs 788, 789, 797 to 802 (with reference to recital 499 of the contested decision), 825, 832 and 838 to 848 of the judgment under appeal.

116 Paragraphs 841 and 842 of the judgment under appeal, with reference to recitals 504 and 505 of the contested decision.

117 Paragraphs 843 to 845 (with reference to paragraphs 825 and 832) of the judgment under appeal.

118 Paragraph 846 of the judgment under appeal, with reference to recitals 230 and 276 of the contested decision.

119 Paragraph 848 of the judgment under appeal, with reference to recital 496 of the contested decision.

120 Paragraph 849 et seq., in particular paragraphs 854 to 861 of the judgment under appeal. That also applies to certain complaints concerning the evidence relied on by the Commission with regard to data traffic and the number of online search queries; see paragraph 866 et seq. of the judgment under appeal.

121 Paragraphs 880 to 883 (with reference to paragraphs 650, 656, 657, 665, 671, 678 to 681 and 859) and paragraph 902 (with reference to paragraph 882) of the judgment under appeal.

122 Paragraph 902 (with reference to paragraph 859) of the judgment under appeal.

123 Paragraph 903 (with reference to paragraph 861) of the judgment under appeal. That part of the judgment under appeal does not form part of the appeal.

124 Paragraph 929 et seq. and paragraph 955 et seq. of the judgment under appeal.

125 Paragraph 943 (with reference to paragraph 843) of the judgment under appeal.

126 Paragraph 944 of the judgment under appeal.

127 Paragraphs 929 to 953 of the judgment under appeal.

128 Paragraphs 945 to 953 of the judgment under appeal, with reference to, inter alia, recital 574 of the contested decision.

129 Paragraph 962 et seq. of the judgment under appeal.

130 Paragraphs 978 to 981 (with reference to paragraphs 650, 656, 657, 665, 671, 678 to 681, 859, 880 and 881) of the judgment under appeal.

131 Paragraph 992 (with reference to paragraphs 967 and 980) of the judgment under appeal.

132 Paragraph 683 et seq. (with regard to the exclusivity clause), paragraph 884 et seq. (with regard to the placement clause) and paragraph 982 et seq. (with regard to the prior authorisation clause) of the judgment under appeal.

133 See, with regard to those requirements, judgment of 27 June 2024, Commission v Servier and Others (C‑176/19 P, EU:C:2024:549, paragraph 423 and the case-law cited).

134 Recital 348 of the contested decision contains a list of GSAs containing the exclusivity clause concluded with 34 direct partners which typically included all their websites. Recital 349 of, and footnote 464 to, that decision mention 69 other such partners which were unable to ascertain whether they had included all their websites in those GSAs.

135 See recital 364 et seq. of the contested decision.

136 Recital 388 of the contested decision.

137 Recital 277 et seq. of the contested decision.

138 Recital 389 of the contested decision.

139 The respective shares of the web domains concerned for those countries are broken down in detail in a Table 16; see recital 390 of the contested decision.

140 Recital 394, read in conjunction with recital 382, of the contested decision.

141 See recital 382 of the contested decision and Table 13 mentioned therein.

142 Recital 395, with reference to recitals 381 to 383 of the contested decision.

143 In that regard, the Commission referred to the judgment of 1 April 1993, BPB Industries and British Gypsum v Commission (T‑65/89, EU:T:1993:31, paragraph 73, according to which ‘the right to terminate a contract in no way prevents its actual application until such time as the right to terminate it has been exercised’).

144 See recital 496 of the contested decision.

145 Recital 498 et seq. of the contested decision.

146 Recital 511 et seq. of the contested decision.

147 Recitals 512 and 513 and Tables 24, 25 and 26 of the contested decision.

148 Recital 515, read in conjunction with recital 390, of the contested decision.

149 Recitals 516 and 517 of the contested decision.

150 Recital 518 and Table 27 of the contested decision.

151 Recital 520, with reference to recital 276 of the contested decision.

152 Recital 526 of the contested decision.

153 Recital 527, with reference to Section 7.3.1 and recital 519 of the contested decision.

154 See recital 574 of the contested decision.

155 Recitals 577 to 579 of the contested decision.

156 Recital 584 et seq. of the contested decision.

157 Recital 586 and Table 28 of the contested decision.

158 Recital 587 and Table 29 of the contested decision.

159 Recital 588, with reference to recital 515 of the contested decision.

160 Recital 589, with reference to recital 518 of the contested decision.

161 Recital 591, with reference to recital 276 of the contested decision.

162 Point 2 of recital 592 of the contested decision.

163 Paragraphs 717 to 727 (exclusivity clause), 893 to 904 (placement clause), and 988 to 1002 (prior authorisation clause) of the judgment under appeal.

164 Paragraphs 698 to 700 and 710 et seq. of the judgment under appeal.

165 Paragraph 703 of the judgment under appeal.

166 Paragraphs 717 and 722 of the judgment under appeal.

167 Paragraph 718 of the judgment under appeal.

168 Paragraphs 893 to 895 of the judgment under appeal.

169 Paragraphs 896 to 898 of the judgment under appeal.

170 Paragraphs 903 and 904 of the judgment under appeal.

171 Paragraphs 988 and 989 (with reference to paragraphs 892 and 893) of the judgment under appeal.

172 Paragraph 990 (with reference to paragraph 897) of the judgment under appeal.

173 See, to that effect, paragraph 991 (with reference to paragraph 898) of the judgment under appeal.

174 Paragraphs 993 (with reference to paragraphs 969 and 991) and 994 of the judgment under appeal.

175 See, to that effect, order of 28 September 2006, Unilever Bestfoods v Commission (C‑552/03 P, EU:C:2006:607, paragraph 55, according to which the Court of First Instance ‘did not err in law when it held in paragraph 105 of the judgment under appeal that, since the possibility of terminating the distribution agreements does not in any way preclude the effective enforcement of those agreements during the period in which that option is not used, it is necessary to take into account the actual duration of those agreements in assessing their effects on the relevant market’).

176 With regard to the consideration given to the ‘status quo bias’ for pre-installed apps on an operating system, see judgment of 2 July 2026, Google and Alphabet v Commission (C‑738/22 P, EU:C:2026:533, paragraph 210 et seq.). See also, in that regard, my Opinion in Google and Alphabet v Commission (C‑738/22 P, EU:C:2025:463, point 95 et seq.).

177 For example, the use of the expressions ‘average duration’ and ‘long’ in recitals 388, 398, 519, 525 and 590 of the contested decision is unclear.

178 In particular recitals 388, 398 and 399 of the contested decision.

179 In particular recitals 519, 525 and 526 of the contested decision.

180 In particular recitals 590 and 594 of the contested decision.

181 Paragraphs 893 to 895 of the judgment under appeal.

182 See, instead, the extensive list of those GSAs in recitals 348 and 349 of the contested decision. Google merely claims in that regard that their duration was neither explicitly mentioned nor assessed in the decision.

183 Paragraph 717 of the judgment under appeal (emphasis added).

184 See paragraph 703 of the judgment under appeal.

185 At the hearing, Google denied having evidence to that effect and merely referred in that regard to paragraph 71 of the response (see also, however, paragraphs 670 and 671 of the judgment under appeal), according to which it had submitted during the administrative procedure inter alia that direct partners conducted competitive processes and often negotiated simultaneously with both Google and Microsoft/Yahoo! throughout the relevant period. However, in 2015 Microsoft had successfully concluded a long-term contract with AOL. The submission by Surfboard reproduced in paragraph 686 of the judgment under appeal also confirms that the GSAs with Google were very often renewed.

186 See the summary above, points 85 to 99.

187 See recital 228 et seq. and recital 274 et seq. of the contested decision. In 2016 those shares were well above [ confidential ] % on most national markets for online search advertising and [ confidential ] % on the market for online search advertising intermediation in the EEA (recital 253) and, in the latter case, had increased particularly sharply since 2006 (with a market share still above [ confidential ] %) (recital 280 et seq.).

188 Based on the gross revenues generated by Google from its intermediation services and the number of search queries concerned and web domains and websites visited; see above points 107 and 109.

189 Paragraph 683 et seq. (with regard to the exclusivity clause), paragraph 884 et seq. (with regard to the placement clause) and paragraph 982 et seq. (with regard to the prior authorisation clause) of the judgment under appeal.

190 See above, points 86 to 99.

191 See above, points 46, 51 and 56.

192 Point 41 et seq., point 86 and point 116 et seq.

193 See, inter alia, judgments of 10 September 2024, Google and Alphabet v Commission (Google Shopping) (C‑48/22 P, EU:C:2024:726, paragraph 166), and of 2 July 2026, Google and Alphabet v Commission (C‑738/22 P, EU:C:2026:533, paragraph 188: ‘on the basis of specific, tangible points of analysis and evidence’).

194 With regard to the aim of the obligation to state reasons incumbent on the General Court, which is inter alia to enable the Court of Justice to exercise its power of review, see judgments of 26 September 2024, Orlen v Commission (C‑255/22 P, EU:C:2024:790, paragraph 171), and of 26 February 2026, Cargolux Airlines v Commission (C‑401/22 P, EU:C:2026:132, paragraph 151 and the case-law cited).

195 They relate to paragraphs 683, 698, 699 and 702 to 714 of the judgment under appeal.

196 These alone account for 40 paragraphs in the Commission’s appeal.

197 See, in particular, paragraphs 704 to 709 of the judgment under appeal.

198 See judgment of 24 October 2024, Commission v Intel Corporation (C‑240/22 P, EU:C:2024:915, paragraphs 328 and 329).

199 Paragraph 890 et seq. of the judgment under appeal.

200 Paragraphs 840, 859, 874, 882, 883, 885, 902 and 967 of the judgment under appeal.

201 Judgment of 24 October 2024, Commission v Intel Corporation (C‑240/22 P, EU:C:2024:915).

202 They relate to paragraph 892 et seq. of the judgment under appeal.

203 Paragraph 986 et seq. of the judgment under appeal.

204 Paragraphs 944, 967 and 980 et seq. of the judgment under appeal.

205 They relate to paragraph 989 et seq. of the judgment under appeal.

206 Recitals 625 to 630 of the contested decision.

207 See paragraphs 1003 and 1004 of the judgment under appeal.

208 Judgment of 2 July 2026, Google and Alphabet v Commission (C‑738/22 P, EU:C:2026:533, paragraphs 381 and 382 and the case-law cited).

209 See, to that effect, judgment of 2 July 2026, Google and Alphabet v Commission (C‑738/22 P, EU:C:2026:533, paragraphs 385 to 387; emphasis added).

210 See, to that effect, judgment of 2 July 2026, Google and Alphabet v Commission (C‑738/22 P, EU:C:2026:533, paragraph 398 and the case-law cited).

211 Judgments of 7 September 2016, Pilkington Group and Others v Commission (C‑101/15 P, EU:C:2016:631, paragraph 73); of 27 April 2017, FSL and Others v Commission (C‑469/15 P, EU:C:2017:308, paragraph 78); and of 18 March 2021, Pometon v Commission (C‑440/19 P, EU:C:2021:214, paragraph 137).

212 See paragraphs 189 to 210 of the application at first instance.

Перевод на русский: GigaChat-2, 13.11.2026. Машинный перевод, вычитывается редакцией.