ECLI:EU:C:2026:861

Full text

OPINION OF ADVOCATE GENERAL

BIONDI

delivered on 15 October 2026 (1)

Case C‑627/24 P

Bytedance Ltd

European Commission

( Appeal – Regulation (EU) 2022/1925 – Designation of gatekeepers – Online social networking service – Decision of the European Commission designating Bytedance as a gatekeeper in respect of its online social networking service TikTok – Action for annulment )

I. Introduction

1. By its appeal, Bytedance Ltd seeks to have set aside the judgment of the General Court in Bytedance v Commission. (2) In that judgment, the General Court upheld the European Commission’s decision to designate that undertaking, pursuant to the DMA, (3) as a gatekeeper with respect to its digital platform service, TikTok. As this is the first dispute relating to the DMA to reach the Court of Justice, the present proceedings thus provide an opportunity to clarify how a central component of the framework established by the DMA operates, namely the designation mechanism.

2. That mechanism is of particular importance as the designation of an undertaking as a gatekeeper by the Commission triggers the application of the obligations set out in the DMA. Moreover, where the quantitative thresholds laid down in Article 3(2) of the regulation are met, the undertaking concerned is obliged to notify the Commission accordingly. By the operation of the presumptions based on those thresholds, the undertaking is deemed a gatekeeper and bears the burden of adducing arguments and evidence to challenge those presumptions. By and large, the DMA rests on an ex ante designation mechanism and is, in that respect, prospective in nature. That said, although the DMA establishes a new legal framework, the approach it embodies is not without precedent in EU law. Its novelty lies rather in bringing a familiar legislative technique to the regulation of the digital sector. The Court’s task is not to revisit the policy choices underlying the DMA, but to give effect to those choices through the proper interpretation and application of that regulation. That understanding shall guide the considerations that follow.

II. Background to the dispute, the decision at issue, the judgment under appeal and the proceedings before the Court of Justice

3. Bytedance operates, together with the companies which it controls directly or indirectly (collectively also referred to as ‘Bytedance’ when referencing the undertaking designated as a gatekeeper), inter alia, the digital platform TikTok. In the European Union, that platform was launched in 2018.

4. On 3 July 2023, the applicant submitted a notification to the Commission in which it maintained, inter alia, that TikTok was a video-sharing platform. The applicant argued that Bytedance cannot be presumed to be a gatekeeper and that, in any event, the presumptions laid down in Article 3(2) of the DMA for the purpose of designating an undertaking as a gatekeeper were rebutted on the basis of the arguments and evidence which it had submitted.

5. By letter of 26 July 2023, the Commission informed the applicant of its preliminary views concerning TikTok’s classification as an online social networking service (‘OSN’) and Bytedance’s designation as a gatekeeper with respect to that service. By letter of 2 August 2023, the applicant replied to those preliminary views.

6. By a decision of 5 September 2023 (‘the decision at issue’), the Commission found that TikTok was an OSN and that Bytedance met the thresholds laid down in Article 3(2) of the DMA, so that it could be presumed to be a gatekeeper. The Commission also considered that the arguments presented by the applicant to rebut the presumptions were not sufficiently substantiated so as manifestly to call them into question and rejected its arguments without opening a market investigation. Consequently, the Commission designated Bytedance as a gatekeeper with respect to TikTok.

7. By the judgment under appeal, the General Court found no error in the Commission’s assessment of the arguments submitted by Bytedance to challenge the presumptions. In that regard, the General Court found that the standard of proof applied by the Commission was correct. It also rejected the plea alleging a breach of the right to be heard. Consequently, the General Court dismissed Bytedance’s action in its entirety.

8. By its appeal, Bytedance claims that the Court should (i) set aside the judgment under appeal, (ii) annul the decision at issue and (iii) order the Commission to pay its own costs as well as Bytedance’s costs in connection with the present proceedings and the proceedings before the General Court. The Commission contends that the Court of Justice should (i) dismiss the appeal and (ii) order Bytedance to pay the costs.

9. Bytedance and the Commission presented oral argument at the hearing held before the Court of Justice on 12 May 2026.

III. Analysis

10. In support of its appeal, Bytedance raises two grounds of appeal. The first asserts an infringement of Article 3(1) and (5) of the DMA and is composed of five parts. The second ground of appeal alleges a breach of Bytedance’s right to be heard. In its response, the Commission formulates a request for the substitution of grounds concerning, inter alia, that right.

11. At the Court’s invitation, the present Opinion will focus exclusively on the first and third parts of the first ground of appeal. It will also address the arguments relating to the Commission’s application for substitution of grounds.

12. To place my analysis in context, I shall begin by making some preliminary observations on the DMA’s presumption-based designation mechanism (Section A). I will then examine the first part of the first ground of appeal concerning the ‘standard of proof’ required under the DMA to rebut the presumptions established in Article 3(2) of that regulation (section B). I will subsequently turn to discuss the third part of the that ground of appeal, which centres on the presumption set out in Article 3(2)(b) of the DMA that a service is an important gateway (section C). Finally, I will consider the requirements flowing from the rights of defence in the context of a presumption-based designation procedure (section D).

A. The presumption-based designation mechanism within the DMA framework

1. The qualitative requirements for designation as a gatekeeper and the role of presumptions based on meeting the quantitative thresholds

13. The substantive test for the designation of an undertaking as a gatekeeper is based on three cumulative and qualitative requirements set out in Article 3(1) of the DMA, namely that (a) it has a significant impact on the internal market, (b) it provides a particular digital platform service, known under the regulation as a core platform service (CPS), which serves as an important gateway for business users to reach end users, and (c) it enjoys an entrenched and durable position, in its operations, or it is foreseeable that it will enjoy such a position in the near future.

14. These qualitative requirements are accompanied by three corresponding quantitative thresholds on which the presumptions laid down in Article 3(2) of the DMA are based. The EU legislature must have considered those thresholds to be proxies for the qualitative requirements set out in Article 3(1) of the DMA.

15. Under Article 3(3) of the DMA, an undertaking meeting the quantitative thresholds is under an obligation to notify the Commission and provide it with relevant information related to those thresholds. In accordance with Article 3(4) of the DMA, upon receipt of the complete information, the Commission is to designate an undertaking meeting the quantitative thresholds as a gatekeeper without undue delay and at the latest within 45 working days of receiving that information.

16. Regardless of the presumption-based designation mechanism, Article 3(8) of the DMA empowers the Commission to open a market investigation to ascertain whether an undertaking that does not meet the quantitative thresholds nevertheless satisfies the qualitative requirements under Article 3(1) of the DMA. In such a case, the Commission is unable to rely on presumptions and bears the burden of proving that an undertaking meets those requirements. In its decision-making practice to date, the Commission has generally dealt with scenarios where the presumption-triggering thresholds are met. That is also the case in the present proceedings.

17. The presumptions laid down in paragraph 2 of Article 3 of the DMA are not conclusive, and paragraph 5 thereof sets out a procedure for their rebuttal.

18. The first subparagraph of Article 3(5) of the DMA states that an undertaking providing CPSs may present, with its notification, sufficiently substantiated arguments to demonstrate that, exceptionally, although it meets all the thresholds in paragraph 2 of that article, due to the circumstances in which the relevant CPS operates, it does not satisfy the requirements listed in paragraph 1 thereof.

19. The second subparagraph of Article 3(5) of the DMA provides that, where the Commission considers that the arguments submitted pursuant to the first subparagraph by the undertaking are not sufficiently substantiated because they do not manifestly call into question the presumptions, it may reject those arguments without opening a market investigation pursuant to Article 17(3) of that regulation.

20. The third subparagraph of Article 3(5) states that where the undertaking providing CPSs does present such sufficiently substantiated arguments manifestly calling into question the presumptions laid down in Article 3(2), the Commission may, within the prescribed time limit, open a market investigation pursuant to that Article 17(3).

21. The threshold issue for the purposes of the appeal is therefore whether Article 3(5) of the DMA imposes an obligation on the Commission to open a market investigation where an undertaking submits arguments manifestly calling into question the presumptions laid down in Article 3(2) of the regulation.

2. The decision to open a market investigation

22. The exchanges at the hearing suggested that the wording of the third subparagraph of Article 3(5) of the DMA (‘the Commission may … open [a market investigation]’) supports an interpretation according to which that provision confers on the Commission a discretion as to whether to open such an investigation. That is not, however, the only way in which that provision can be understood. In my view, the use of the term ‘may’ does not, in that context, confer a discretionary power on the Commission, rather it identifies the procedural steps available to the Commission once the relevant conditions laid down in that provision are satisfied.

23. On that interpretation, where an undertaking submits sufficiently substantiated arguments manifestly calling into question the presumptions laid down in Article 3(2) of the DMA, the Commission cannot reject those arguments without opening a market investigation, unless it is already apparent that the requirements under Article 3(1) are not satisfied in view of the circumstances in which the CPS operates. By contrast, where the arguments submitted do not manifestly call into question those presumptions, the Commission may reject them without opening a market investigation.

24. Such a reading is confirmed, first, by a systemic interpretation of the DMA.

25. The mechanism laid down in Article 3(5) of the DMA corresponds to the Commission’s powers set out in Article 17(3) of that regulation which relates to market investigations. In particular, the first sentence of the latter provision provides that where the undertaking providing CPSs satisfies the thresholds set out in Article 3(2) of the DMA, but has presented sufficiently substantiated arguments that manifestly call into question the presumption from Article 3(2), namely the hypothesis of the third subparagraph of Article 3(5), the Commission is to endeavour to conclude the market investigation within a time limit specified by the DMA. Although Article 17(3) expressly governs the conclusion of the procedure, it necessarily implies an obligation on the Commission to open that procedure where the undertaking has presented sufficiently substantiated arguments manifestly calling into question the presumptions.

26. Admittedly, Article 17(3) of the DMA could be understood as merely prescribing the time limit within which the Commission is to conclude a market investigation. Nonetheless, such an interpretation would largely deprive that article, entitled ‘Market investigation for designating gatekeepers’, of its independent function. Moreover, in contrast with paragraph 1 thereof, which provides that ‘the Commission may conduct a market investigation for the purpose of examining whether an undertaking providing core platform services should be designated as a gatekeeper pursuant to Article 3(8)’, paragraph 3 thereof, referred to in Article 3(5) of that regulation, appears to leave no discretion to the Commission as to its decision to open a market investigation. Thus, if anything, it is Article 17(1) and (3) rather than Article 3(5) and (8) that determines the Commission’s margin of discretion.

27. Secondly, the fourth subparagraph of Article 3(5) of the DMA also supports that interpretation. The second and third subparagraphs govern two distinct situations: where the undertaking does not present sufficiently substantiated arguments manifestly calling into question the presumptions (second subparagraph) and where the undertaking does succeed in doing so (third subparagraph), respectively. (4) The fourth subparagraph sets out the logical consequence of the third subparagraph and of the mandatory opening of a market investigation. Under that fourth subparagraph, ‘if the Commission concludes that the undertaking … was not able to demonstrate that the relevant [CPSs] do not satisfy the requirements of paragraph 1 of this Article, it shall designate that undertaking as a gatekeeper in accordance with the procedure laid down in Article 17(3).’ The Commission’s conclusion to which the fourth subparagraph refers is reached following a market investigation opened in accordance with the third subparagraph. Consequently, where an undertaking has presented sufficiently substantiated arguments manifestly calling into question the presumptions, the Commission has no alternative if it wishes to pursue designation: it has to follow the market investigation procedure laid down in Article 17 of the DMA. The fourth subparagraph of Article 3(5) simply confirms that obligation.

28. Thirdly, the DMA establishes a framework which ensures a fair balance between the need for a swift designation process and proper consideration of an undertaking’s arguments calling the presumptions into question. On the one hand, the Commission may rely on the quantitative thresholds from Article 3(2), enabling a fast designation process based on the presumptions that the qualitative requirements from Article 3(1) are satisfied. (5) On the other hand, where an undertaking succeeds in manifestly calling those presumptions into question, it is entitled to expect the Commission to undertake further verification of its arguments concerning the non-fulfilment of the qualitative requirements under Article 3(1). Against that background, the concept of ‘sufficiently substantiated arguments manifestly calling into question the presumptions’ lies at the heart of the mechanism of Article 3(5) and has to bear legal significance. Were it otherwise, there would be little practical value in providing an undertaking with the possibility of submitting arguments calling into question the presumptions if, notwithstanding their successful demonstration, the legal consequences remained entirely a matter for the Commission’s discretion.

29. Fourthly, that interpretation is supported by the positions expressed by Bytedance and the Commission at the hearing. I would also point out that it might be appropriate and useful to have regard to the Commission’s own decision-making practice. (6) The Commission’s decisions taken so far confirm the existence of two distinct procedural scenarios. Where an undertaking submits arguments that manifestly call into question the presumptions laid down in Article 3(2) of the DMA, the Commission is required to open a market investigation. By contrast, where the undertaking demonstrates clearly and comprehensively (7) that those requirements are not met – or, as expressed in recent decisions, provides a ‘clear demonstration’ to that effect, (8) the Commission does not initiate the Article 17 procedure and may proceed directly to non-designation.

30. To summarise, if the quantitative thresholds under Article 3(2) of the DMA are met and the undertaking seeks to challenge the resulting presumptions, the specific trigger for opening a market investigation is the presentation of sufficiently substantiated arguments that manifestly call into question those presumptions. Presented with such arguments, the Commission either (i) opens a market investigation to verify whether the undertaking has demonstrated that the qualitative requirements are, exceptionally, not met due to the circumstances in which the relevant CPS operates, or (ii) considers that, already at this stage, it is clear that that is the case; ‘exceptionality’ is thus clearly demonstrated ab initio. Conversely, if the arguments do not manifestly call into question the presumptions, the Commission may reject them without further examination within a market investigation.

31. In that regard, some uncertainty remains as to the point at which the presumptions set out in Article 3(2) of the DMA may be considered to have been manifestly called into question. The issue hinges on the standard of proof required under Article 3(5) of that regulation.

3. The standard of proof required by Article 3(5) of the DMA

32. Although the DMA does not define precisely the standard of proof required to rebut the presumptions under Article 3(2) thereof, it does provide multiple indications in that regard.

33. First of all, the arguments manifestly calling into question the presumptions need to be ‘sufficiently substantiated’. Recital 23 of the DMA states, in its second sentence, that ‘the burden of adducing evidence that the presumption deriving from the fulfilment of the quantitative thresholds should not apply should be borne by that undertaking’ and adds, in the third sentence, that the Commission conducts the assessment of ‘the evidence and arguments produced’. Thus, Article 3(5) of the DMA requires more than mere speculation or bare assertions.

34. Next, the standard of proof required by Article 3(5) of the DMA is not satisfied by mere prima facie evidence. Rather, the requirement that the presumptions be ‘manifestly’ called into question indicates that the undertaking must adduce evidence that the presumptions should not apply. If the mere casting of doubt or contesting of the presumptions were to be considered sufficient, the Commission would be obliged to open a market investigation in each case where an allegation to that effect is raised. Such an interpretation would compromise the objective of the presumption mechanism, which is intended to ensure a fast designation procedure and the effective application of the DMA. (9) Moreover, it would also obfuscate the difference between a presumption-based designation pursuant to Article 3(2) thereof and a designation pursuant to Article 3(8) thereof: in practice, in both scenarios, the Commission would ultimately always be required to undertake a market investigation.

35. Lastly, attention must be given to the related question of the way in which the legal presumption operates under the DMA. The undertaking has to succeed in showing that, even though a quantitative threshold under Article 3(2) of that regulation is met, the presumed circumstance (namely, the fulfilment of the corresponding qualitative requirement under Article 3(1) thereof) requires further investigation, since the inference of that circumstance from the quantitative threshold is questionable. In that setting, a presumption can be manifestly called into question only if the logic of the rebutting argument is clear and highly compelling.

36. Thus, in order to ensure the effective application of the DMA and the swiftness of the designation process, (10) as well as to make it possible for its provisions to apply to the market actors whose CPS may impact the digital sector in a manner that the DMA aims to limit, the standard of proof required by Article 3(5) amounts to a high degree of plausibility. (11) The arguments have to be genuinely compelling, with no easily identifiable weaknesses in their capacity to rebut the presumption, in order to be considered as ‘manifestly’ calling it into question.

37. It is in the light of those observations that the grounds of appeal and the Commission’s request should be examined.

B. The first part of the first ground of appeal – standard of proof required by the DMA to challenge the presumptions

38. By the first part of the first ground of appeal, Bytedance challenges paragraphs 39 to 72 of the judgment under appeal and submits that the General Court erred in law by applying an excessively high standard of proof for challenging the presumptions laid down in Article 3(2) of the DMA and for triggering the opening of a market investigation.

39. Bytedance submits that the General Court incorrectly stated in paragraphs 70 and 71 of the judgment under appeal that, under Article 3(5) of the DMA, a market investigation may be opened where the arguments of the undertaking concerned are sufficiently substantiated so as to, exceptionally, manifestly call into question the presumptions. Bytedance underscores that the term ‘exceptionally’ is used solely in the first subparagraph of Article 3(5) of the DMA, and not in the second or third subparagraphs thereof.

40. Bytedance essentially argues that, contrary to the General Court’s position on the matter, the standard of proof required by Article 3(5) of the DMA does not involve any form of an ‘exceptionality’ requirement (addressed in section 1 below). Bytedance also claims that the alleged error permeates the entire judgment under appeal, since the General Court called for it to demonstrate that the requirements set out in Article 3(1) of the DMA are not satisfied, whereas the correct test is whether the sufficiently substantiated arguments manifestly call into question the presumptions from Article 3(2) of the DMA (addressed in section 2 below).

1. The standard of proof applied in the judgment under appeal

41. As I have observed in points 23 to 30 of the present Opinion, in the presumption-based designation mechanism, the ‘exceptionality’ does not concern the Commission’s initial finding that the presumptions are manifestly called into question. Instead, the ‘exceptionality’ concerns the subsequent finding, made by the Commission through a market investigation or ab initio without opening such an investigation, that the requirements set out in Article 3(1) of the DMA are not satisfied despite the quantitative thresholds being met.

42. In that regard, paragraphs 70 and 71 of the judgment under appeal, read in isolation, might give the impression that, in the General Court’s view, the DMA applies an ‘exceptionality’ requirement to the rebuttal of presumptions. The General Court stated, in paragraph 70, that ‘under [Article 3(5) of the DMA,] a market investigation may be opened where the arguments of the undertaking concerned are sufficiently substantiated so as, exceptionally, manifestly to call into question the presumptions laid down in Article 3(2) of that regulation’. In paragraph 71 it made reference both to the term ‘manifestly’ and to the term ‘exceptionally’, and, taken at face value, appeared to rely on them to determine the standard of proof required by the DMA.

43. While the use of the term ‘exceptionally’ in paragraphs 70 and 71 of the judgment under appeal may be unfortunate, a comprehensive reading of the judgment does not disclose the error of law alleged by Bytedance.

44. Before the General Court, Bytedance did not contend that the Commission considered that that undertaking could only ‘exceptionally’ rebut the presumptions. As is clear from paragraph 69 of the judgment under appeal, Bytedance maintained that the standard of proof required by the Commission was too demanding: the Commission required ‘convincing’ evidence, while for Bytedance ‘the standard of proof’ that is sufficient to trigger a market investigation amounts to demonstrating the existence of ‘doubts’ or ‘prima facie’ evidence.

45. In paragraphs 70 and 71 of the judgment under appeal, the General Court was directly responding to that argument made by Bytedance. The General Court’s response is set out in the second part of the latter paragraph, in which it stated that ‘the arguments presented by that undertaking must be capable of showing, with a high degree of plausibility, that the presumptions laid down in Article 3(2) of the DMA are called into question’. The requirement of a high degree of plausibility corresponds to the standard of proof required under the DMA and discussed in point 36 of the present Opinion.

46. That reading of the judgment under appeal is confirmed, moreover, first, by paragraph 60, where the General Court faithfully reproduces Article 3(5) of the DMA by stating that ‘the first subparagraph of paragraph 5 of that article provides that the undertaking concerned may present sufficiently substantiated arguments to demonstrate that, exceptionally, although it meets all the thresholds in paragraph 2 of that article, it does not satisfy the requirements listed in paragraph 1 thereof’. The General Court concludes that ‘where that undertaking does present such sufficiently substantiated arguments, manifestly calling into question those presumptions, the Commission may open a market investigation’. No reference to ‘exceptionality’ is to be found in that conclusion. Furthermore, in paragraph 61 of the judgment under appeal, the General Court explained – again, with no reference to ‘exceptionality’ – that the standard of proof determined by the EU legislature requires the undertaking concerned to submit sufficiently substantiated arguments, manifestly calling into question those presumptions.

47. Secondly, in paragraph 65 of the judgment under appeal, the General Court observed that the Commission asserted that it had applied the standard of proof set out in Article 3(5) of the DMA and described in paragraph 60 of the judgment.

48. Thirdly, in paragraph 66, the General Court even notes that the standard of proof required by the DMA cannot be understood ‘as meaning that the applicant must adduce evidence which rebuts that presumption once and for all’. Consequently, and in contrast to the argument invoked in Bytedance’s reply, the General Court did not require the undertaking exhaustively to disprove the qualitative requirements from Article 3(1) of the DMA.

49. In that setting, in spite of the wording of paragraphs 70 and 71 of the judgment under appeal and the use of the term ‘exceptionally’ contained therein, the standard of proof formulated by the General Court in paragraphs 39 to 72 perfectly matches the standard of proof required by the DMA and described in point 36 of the present Opinion.

2. Application of the standard of proof

50. In my view, Bytedance’s arguments that the judgment under appeal is permeated by an exceptionality requirement also do not support the conclusion that the General Court applied a more demanding standard of proof than the one required by the DMA.

51. In that respect, Bytedance cites (i) paragraph 106 of the judgment under appeal, which it reads as requiring it to ‘demonstrate’ that its Asian businesses faced cultural and regulatory barriers in the European Union; (ii) paragraph 113 thereof, examining whether the undertaking’s low EU turnover could ‘rebut the presumption’ laid down in Article 3(2)(a) of the DMA; (iii) paragraph 116 thereof, requiring it to ‘disprove’ certain circumstances; (iv) paragraph 161 expecting Bytedance to ‘prove’ that the lack of an ecosystem means it is not an important gateway to displace the presumption laid down in Article 3(2)(b) of the DMA; and (v) paragraph 200 thereof, stating that Bytedance must demonstrate that the asymmetric nature of multi-homing resulted from the absence of lock-in effects.

52. In essence, Bytedance’s arguments appear to hinge on the observation that, in various paragraphs of the judgment under appeal, the General Court used the terms ‘demonstrate’, ‘prove’ or ‘disprove’ to uphold the Commission’s conclusion that the presumptions were not manifestly called into question by the arguments submitted.

53. Nonetheless, first, as the Commission pointed out in its response, those terms reflect the fact that the undertaking bears the burden of adducing evidence that the presumptions should not apply. Secondly, the third sentence of recital 23 of the DMA indicates that it is the Commission that conducts the assessment of ‘the evidence and arguments produced’ by the undertaking concerned. The linguistic choices of the General Court (‘demonstrate’, ‘prove’ and ‘disprove’) are natural in the context of that assessment and, while acknowledging that the burden rests on the undertaking, are otherwise neutral in regard to the standard of proof required. Thirdly, the terms ‘demonstrate’, ‘prove’ and ‘disprove’ form part of the language chosen by the General Court to assess whether Bytedance succeeded in manifestly calling into question the presumptions. More specifically, the General Court did not require Bytedance to disprove the presumed circumstances (namely that a given qualitative requirement from Article 3(1) of the DMA is not fulfilled), but to demonstrate or prove that the inference of that circumstance from the quantitative threshold is manifestly questionable. That approach corresponds to the standard of proof required under the DMA. (12)

54. In its reply, Bytedance criticises the judgment under appeal by submitting also that, due to the General Court’s understanding of the standard of proof demanded by the DMA, Bytedance was required to ‘anticipatorily disprove unpredictable predicates for designation’. To illustrate that point, Bytedance argues that, contrary to what the General Court held, by demonstrating that almost all of its users multi-home across several platforms, it succeeded in showing that there is no automatic link between the basis of the presumption laid down in Article 3(2)(b) of the DMA (a high number of users) and the presumed circumstance from Article 3(1)(b) (the CPS in question is an important gateway that businesses depend on to reach consumers).

55. However, that line of argument disregards the manner in which the legal presumptions operate against the undertaking meeting one of the quantitative thresholds of Article 3(2) of the DMA. A case in point is provided by Bytedance’s own illustration, wherein the presumed circumstance was inferred from the number of users and not from the fact that they do not engage in multi-homing. The demonstration of prevalent multi-homing alone is therefore not sufficient manifestly to call into question the presumption because the lack of multi-homing is only one of the proxies for the presumed important gateway position of the CPS in question.

56. All things considered, in spite of the wording of paragraphs 70 and 71 of the judgment under appeal and the use of the term ‘exceptionally’ contained therein, the General Court benchmarked the Commission’s assessment of the arguments and evidence submitted by Bytedance against the standard of proof required under the DMA and described in point 36 of the present Opinion.

57. In my view, and against that background, neither the arguments invoked against the abstract formulation of the standard of proof in the judgment under appeal, nor its concrete applications, amount to an error of law as alleged by Bytedance in its appeal. It follows from the foregoing that the first part of the first ground of appeal should be rejected.

C. The third part of the first ground of appeal – the important gateway requirement and its corresponding presumption

58. In the third part of the first ground of appeal, Bytedance challenges paragraphs 132 to 285 of the judgment under appeal and argues that the General Court erred in law by upholding the Commission’s finding that the applicant did not manifestly call into question the presumption laid down in Article 3(2)(b) of the DMA.

59. The third part of the first ground of appeal itself is composed of six complaints, which allege: (1) the misapplication of the criterion of having an ‘ecosystem’; (2) the substitution of the General Court’s own reasoning for that of the Commission in relation to significant network effects; (3) the substitution of the General Court’s own reasoning for that of the Commission’s in relation to multi-homing and users lock-in; (4) the distortion of the evidence and the failure to state reasons as regards multi-homing, intensity of use and interoperability; (5) the misapplication of Article 3(1)(b) of the DMA in relation to relative scale; and (6) the application of the incorrect test in the analysis of whether its service is an ‘important gateway’.

1. The alleged misapplication of the ecosystem criterion

60. As a preliminary remark, in paragraphs 128 to 131 of the judgment under appeal, the General Court observed that while the DMA does not provide a definition of an ‘ecosystem’, the meaning of that concept can be inferred from its recitals. The General Court continued by explaining that an ‘ecosystem’ may consist of one or more CPSs and other services connected to them, for example by means of technological links or interoperability; having such an ecosystem is liable to exacerbate entry barriers for the competitors of those undertakings and to increase the cost of switching providers for end users, making it more difficult for existing or new market operators to compete with those undertakings or challenge their position. Consequently, having a digital platform ecosystem can constitute a relevant factor for the purposes of assessing whether the undertaking concerned is a gatekeeper and whether, more particularly, the CPS in question is an important gateway within the meaning of Article 3(1)(b) of the DMA.

61. Bytedance argues that, in paragraphs 141 to 145 of the judgment under appeal, the General Court incorrectly repeats the Commission’s conclusion that it had not substantiated its claim that it lacks a whole platform ecosystem. Bytedance considers that the General Court should not have relied on that conclusion as the Commission had infringed its rights of defence.

62. However, the infringement alleged by the applicant was addressed in paragraph 353 of the judgment under appeal, in which the General Court found that the Commission had infringed its right to be heard by not giving it the opportunity to submit observations during the administrative procedure on the Commission’s finding that Bytedance operates an ecosystem. The General Court’s finding in that paragraph is the subject of the second ground of appeal. I propose to follow the order in which Bytedance has set out the grounds and arguments in its appeal. Accordingly, I shall address the separate issue of Bytedance’s right to be heard in the context of my analysis of the second ground of appeal.

63. By contrast, and concerning the specific argument raised by Bytedance in the third part of its first ground of appeal, the General Court found that there was no breach of Bytedance’s rights in the Commission’s appraisal that it had not substantiated its argument on its lack of an ecosystem.

64. It is also worth noting that, in recital 132 of the decision at issue, the Commission considered that Bytedance operates its own ecosystem consisting of highly popular video-editing services, enterprise software, advertising, news and healthcare applications. However, in recitals 129 and 133 of that decision, the Commission also argued that, even if an ecosystem is lacking, Bytedance had not succeeded in calling into question the presumption laid down in Article 3(2)(b) of the DMA. In essence, the Commission considered that having an ecosystem is not a required constituent to support a finding that a service is an important gateway for business users to reach end users within the meaning of Article 3(1)(b) of that regulation. Overall, as repeatedly explained in numerous parts of the decision at issue and of the judgment under appeal, an important gateway position within the meaning of the latter provision may equally result from the cumulative effect of other factors, such as significant network effects or lock-in effects of the platform. Bytedance’s unwillingness to accept that line of reasoning remains a recurring theme of its appeal.

65. Bytedance’s remaining arguments are based on an inaccurate reading of the judgment under appeal. First, Bytedance submits that that judgment is contradictory in so far as it acknowledges the relevance of an ecosystem in paragraphs 128 to 132, just to conclude in paragraph 133 that the absence of an ecosystem is not relevant to whether its service is an important gateway. However, paragraph 133 states that the fact that a CPS is not part of an ecosystem is insufficient to demonstrate that the service is not an important gateway, rather than that it is irrelevant in that regard. Secondly, Bytedance argues that the conclusion set out in paragraphs 152 to 154 of the judgment under appeal, which finds that the lack of an ecosystem does not prevent the CPS from having significant networks effects, ‘misses the point’. Bytedance’s submission is that the General Court erred in law in concluding that the lack of an ecosystem is irrelevant simply because barriers to switching may also arise due to network effects. However, the General Court did not find that the lack of an ecosystem is irrelevant, but that it is insufficient to challenge the applicable presumption. Thirdly, Bytedance argues that, in paragraphs 148 to 151 of the judgment under appeal, the General Court incorrectly found that the absence of an ecosystem did not prevent TikTok from growing its user base. Bytedance submits that the mere existence of a large user base in itself says nothing about the legal question relevant from the viewpoint of Article 3(1)(b) of the DMA, namely whether, owing to barriers to switching as a result of the platform’s ecosystem, business users depend on the platform to reach end users. However, the point made by the Commission and upheld by the General Court was that, even in the absence of an ecosystem operated by other undertakings, Bytedance had nevertheless been able to grow its user base to a scale that was not negligible compared with that of those undertakings. Accordingly, the fact that Bytedance did not operate an ecosystem was insufficient to call into question the presumed important gateway position of the CPS arising from the scale and network effects enjoyed by a widespread and commonly used digital service. (13)

66. In the light of the foregoing, I take the view that the first complaint of the third part of the first ground of appeal is unfounded.

2. The alleged substitution of the Commission’s reasoning in relation to significant network effects

67. Recital 2 of the DMA cites ‘very strong network effects’ as an example of a characteristic of CPSs that can be exploited by the undertakings providing them. Recital 13 of the DMA then adds that ‘very strong network effects’ are among the features which render weak contestability and unfair practices in the digital sectors more frequent and more pronounced for CPSs’ exhibiting those features. Recital 32 also stresses that ‘network effects’ limit the contestability of services concerned.

68. While the DMA itself does not define the concept of ‘network effects’ in more detail, in paragraph 153 of the judgment under appeal, the General Court explained that network effects occur when the value of a product or service increases as more people use it. Consequently, online social networks generate strong network effects and data-driven advantages when, gradually, their number of users increases, since the more users there are, the more useful that network is to those users and the more its value increases for them, which, in turn, attracts more users. In addition, the greater the number of end users of an online social network, the more business users are attracted to that network. I would add that significant network effects are therefore one of the CPSs’ characteristics capable of establishing or reinforcing the important gateway position of a platform service within the meaning of Article 3(1)(b) of the DMA.

69. Bytedance alleges that the General Court supplemented the reasoning in the decision at issue with its own findings and misapplied Article 3(1)(b) of the DMA in relation to significant network effects. That complaint is founded on two arguments.

70. First, Bytedance criticises paragraphs 152 to 160 of the judgment under appeal for stating that it had not argued that TikTok lacked significant network and lock-in effects. Bytedance argues that the decision at issue did not dispute that TikTok lacks such effects.

71. However, in particular in paragraph 154 of the judgment under appeal, the General Court did not consider that Bytedance had chosen not to argue that TikTok lacked significant network effects. In fact, the General Court considered that the arguments put forward by Bytedance in that regard were insufficient to establish the lack of those effects. The General Court observed that Bytedance (i) had confined itself to arguing that the alleged lack of an ecosystem leads to the absence of significant network effects, and (ii) had not made any independent argument suggesting that TikTok lacks such effects. Moreover, the argument that the absence of an ecosystem results in an absence of significant network effects not only contradicts the General Court’s interpretation, (14) but also contradicts Bytedance’s own argument that the existence of an ecosystem and the presence of significant network effects constitute two distinct questions.

72. In addition, as regards Bytedance’s submission that the decision at issue did not dispute TikTok’s lack of significant network effects, I would draw attention to the fact that the claims concerning those effects are addressed in recitals 127 to 133 of that decision. (15) The relevant section of the decision at issue is even entitled ‘Alleged lack of an ecosystem and of significant network effects’. In that section, the Commission addressed Bytedance’s arguments and considered that they are not capable of manifestly calling into question the presumption under Article 3(2)(b) of the DMA. Furthermore, it must not be overlooked that it was for Bytedance to challenge that presumption by demonstrating that TikTok does not produce such effects, rather than for the Commission to prove the contrary and explain why it had come to that conclusion.

73. Secondly, Bytedance criticises paragraph 158 of the judgment under appeal for stating that it could not argue that it lacked significant network effects as it did not contest TikTok’s classification as an OSN. Bytedance appears to read the judgment as implying that an OSN necessarily exhibits network effects and any evidence to the contrary is therefore ineffective. However, contrary to Bytedance’s line of argument, TikTok’s gateway position was presumed because the quantitative threshold laid down in Article 3(2)(b) of the DMA was met by the CPS at issue, and not because it was classified as an OSN.

74. In that vein, paragraph 158 of the judgment under appeal continues the reasoning commenced in paragraph 155 thereof, where the General Court addressed Bytedance’s plea that TikTok lacks network effects associated with the ‘social graph’ which characterises ‘real’ social networks such as Facebook or Instagram. In paragraphs 155 to 158, the General Court simply considered that TikTok is a CPS classified as an OSN and, even if it does not rely on a ‘social graph’, Bytedance still had to demonstrate that its service lacked significant network effects in order to call into question the presumption that it was an important gateway.

75. For those reasons, I conclude that the second complaint of the third part of the first ground of appeal is unfounded.

3. The alleged misapplication of Article 3(1)(b) of the DMA in respect of multi-homing and user lock-in and substitution of reasoning

76. Bytedance argues that, in paragraphs 174 to 214 of the judgment under appeal, the General Court misapplied Article 3(1)(b) of the DMA relating to the requirement that a service be an important gateway by rejecting the importance of multi-homing, failing to state reasons in its lock-in analysis and misinterpreting the criterion that business users depend on its service to reach end users.

(a) Multi-homing

77. I shall precede the following considerations by observing that, in paragraph 171 of the judgment under appeal, the General Court noted that ‘multi-homing’ describes the situation in which users use several competing digital services in parallel. It then analysed recitals 2 and 13 of the DMA concluding, in paragraph 174 of the judgment under appeal, that the absence of multi-homing is a relevant factor, among others, capable of characterising a gatekeeper providing a CPS which is an important gateway within the meaning of Article 3(1)(b) of the DMA. I would add that, conversely, the existence of multi-homing could, in some circumstances, be a contributing factor in challenging a presumption that a service is an important gateway.

78. Bytedance claims that, in paragraphs 182 to 185 of the judgment under appeal, the General Court disregarded multi-homing in relation to OSNs. It argues that the General Court stated that multi-homing, ‘even to a significant degree’, is not a relevant criterion for OSNs if it is a common practice. However, I fail to see an error in the General Court’s reasoning.

79. First of all, as a preliminary point, the reasoning of the General Court is based on the interpretation, set out in paragraph 182 of the judgment under appeal, that the relevance of different factors, as elements to be taken into consideration for the purposes of determining whether a CPS is an important gateway, may vary according to the different CPS categories. I see nothing controversial in that finding. The services falling within those categories operate differently and arguments of different sorts might be used to challenge the applicable presumptions. Contrary to the argument put forward by Bytedance, that interpretation is not contradicted by the DMA, which, in Article 3(8), mentions multi-homing as relevant to assess whether a service is an important gateway without distinguishing between CPS categories. In fact, Article 3(8) of the DMA explicitly states that the Commission is to take into account some or all elements mentioned in that provision in so far as they are relevant for the undertaking providing the CPSs under consideration. Thus, the wording of that provision already indicates that it is intended to offer guidance to the Commission and structure its market investigation. It does not, however, set out a mandatory step-by-step checklist.

80. Next, TikTok was categorised as an OSN and – through the operation of presumption laid down in Article 3(2)(b) of the DMA – it is presumed to be an important gateway service. It is against that backdrop that the General Court simply states, to paraphrase paragraphs 182 to 185 of the judgment under appeal: as illustrated by the data presented by Bytedance itself, OSNs in general show a degree of multi-homing and it is for the undertaking concerned to demonstrate that, in spite of that feature, it is manifestly questionable that its service is an important gateway within the meaning of Article 3(1)(b) of the DMA.

81. Finally, I would add that, under the first subparagraph of Article 3(5) of the DMA, in order to rebut a presumption that it meets a requirement listed in Article 3(1), an undertaking has to present arguments demonstrating that, ‘due to the circumstances in which the relevant [CPS] operates’, it does not satisfy the requirements listed in Article 3(1) of the DMA. Thus, the reliance on general or abstract statements about the nature of a category of services is not sufficient to rebut the presumptions. A fortiori, showing that the CPS at issue is characterised by a significant degree of multi-homing is not sufficient manifestly to call into question the presumption if, as illustrated by Bytedance itself in the present case, a given category of CPS generally tends to demonstrate a significant degree of multi-homing.

82. The argument related to multi-homing is therefore unfounded.

(b) Lock-in effects

83. The lock-in effects are mentioned in recitals 2, 13 and 25 as well as Article 3(8) of the DMA as being one of the characteristics of a CPS that meets the requirements laid down in Article 3(1) of the DMA. Point (e) of the second subparagraph of Article 3(8) of the DMA characterises user lock-in as an element that involves switching costs and behavioural bias, which reduces the ability of business users and end users to switch or multi-home. Taken together, all those elements support the following understanding of the concept of lock-in effects: it denotes a situation where users are de facto constrained to remain within a gatekeeper’s service because of high switching costs, behavioural biases or other factors that make switching or multi-homing unlikely. Consequently, the absence of such effects could support an argument seeking to call into question a presumption under Article 3(2)(b) that a service is an important gateway.

84. Bytedance argues that paragraphs 179 and 202 of the judgment under appeal incorrectly state that usage intensity is a criterion to determine user lock-in, and, therefore, indicates an important gateway position. The argument is composed of two parts, namely that the General Court substituted its reasons for those of the Commission and that it misapplied the law in respect of lock-in effects.

(1) The alleged substitution of reasoning

85. Bytedance argues that, in paragraphs 179 and 202 of the judgment under appeal, the General Court substituted its own reasoning for that given by the Commission in recitals 131 to 142 of the decision at issue.

86. In that regard, in paragraph 179 of the judgment under appeal, the General Court stated that lock-in effects may arise even in the presence of multi-homing. The central point of that paragraph, namely the interplay between lock-in effects and multi-homing, was discussed in recitals 136 to 139 of the decision at issue. (16) The judgment under appeal echoes those considerations in direct response to the plea raised by Bytedance before the General Court. (17)

87. However, Bytedance argues that, unlike paragraph 179 of the judgment under appeal, the decision at issue does not state that lock-in effects arise from factors such as significant network effects, stronger engagement of certain user categories and behavioural bias. I consider that Bytedance’s argument is not compelling. Paragraph 179 clearly alludes to the factors which, as stated in the decision at issue, may foster lock-in effects even in presence of multi-homing. User engagement is considered in recital 141 of the decision at issue, (18) while, as the Commission points out, the behavioural bias displayed towards a platform within certain groups of users is referred to in recitals 137 and 138 of that decision, as a matter of asymmetric use of OSNs.

88. Arguably, although the term ‘behavioural bias’ does not appear in the decision at issue, its use had no bearing on the General Court’s reasoning which is consistent with that of the Commission. (19) The substance of paragraph 179 of the judgment under appeal is not different from that of recitals 131 to 142 of the decision at issue: significant multi-homing does not necessarily imply lack of lock-in effects, and users’ preferences may contribute to such effects.

89. I would also add that, even if no reference to behavioural bias were made in the decision at issue, the General Court could not successfully be criticised for having substituted its own assessment for that of the Commission. Point (e) of the second subparagraph of Article 3(8) of the DMA characterises user lock-in as an element that involves switching costs and behavioural bias reducing the ability of business users and end users to switch or multi-home. In paragraph 179, the General Court simply repeated what the law stipulates in order to respond to Bytedance’s argument, based on an incorrect reading of recitals 2 and 3 and of Article 3(8) of the DMA, (20) that multi-homing necessarily means there is a lack of lock-in effects. Article 3(8) of the DMA states that lock-in effects may hinder multi-homing, but it does not indicate that the prevalence of multi-homing implies a lack of lock-in effects. Lock-in effects may also limit ‘migration of focus’ to another platform (namely the continued use of multiple services, accompanied by shifting particular emphasis from one to another), as the Commission notes in recital 138 of the decision at issue.

90. As regards the remaining argument related to the alleged substitution of the grounds in paragraph 202 of the judgment under appeal, Bytedance argues that the decision at issue did not state that it had not provided any submissions on the absence of alleged lock-in effects. However, that decision analyses the alleged multi-homing and absence of lock-in effects in recitals 134 to 142 thereof. The related section of the decision is even entitled ‘Alleged multi-homing and absence of lock-in’. The analysis contained therein demonstrates that the Commission also identified no independent argument on the absence of lock-in effects, but rather considered that Bytedance had been arguing that TikTok lacks such effects as its users engage in multi-homing.

91. I conclude that the argument related to the alleged substitution of grounds with regard to lock-in is therefore unfounded.

(2) The alleged misapplication of law with regard to lock-in effects

92. First, Bytedance claims that paragraphs 179, 189 and 202 of the judgment under appeal misread recital 2 and Article 13 of the DMA, by proceeding from the mistaken legal premiss that lock-in effects and multi-homing need to be assessed separately. Bytedance argues that if users multi-home across multiple platforms they are not, by definition, locked into a (single) platform. Bytedance essentially submits that to challenge the presumption under Article 3(2)(b) of the DMA, the undertaking is required to demonstrate either the lack of lock-in effects or significant multi-homing. However, as observed above in point 89 of the present Opinion, the lack of lock-in effects and multi-homing are not necessarily interconnected and may constitute distinct elements taken into account in the assessment of the important gateway requirement.

93. Secondly, Bytedance criticises the General Court for requiring it to demonstrate low usage intensity as a measure for lock-in effects. It submits that the General Court also failed to state reasons as to why time spent on the platform is the only metric for usage intensity. Bytedance then argues that intensity of use may be best judged by the number of interactions, and not by time spent.

94. In my opinion, however, Bytedance misreads the judgment under appeal. Its criticism concerns paragraph 178, in which the General Court stated that ‘the intensity of that use, that is to say the time spent by a user interacting on that platform, and the importance of such a platform for certain categories of users, may constitute, among other factors, and as the case may be, relevant factors for assessing whether the existence of a certain degree of multi-homing can manifestly call into question the presumption that the CPS concerned is an important gateway’. Thus, the General Court did not consider that the intensity of use is a measure for assessing lock-in effects, but that it is a measure for establishing the importance of extensive multi-homing, which, otherwise, was considered insufficient to call into question that presumption. Moreover, as the Commission points out, the General Court did not hold that time spent is the only metric, but that it is an appropriate one. The General Court also provided an extensive explanation as to why that metric is appropriate, in particular in paragraphs 368 and 369 of the judgment under appeal.

95. Thirdly, and still in the context of intensity of use, Bytedance argues that, contrary to the point made by the General Court in paragraph 188 of the judgment under appeal, the DMA does not suggest that a platform can be an important gateway for a specific subset of end users based on criteria such as age.

96. In my view, Bytedance’s argument is unfounded. The important gateway position is presumed, in accordance with Article 3(2)(b) of the DMA, on the basis of the total number of active end users, while the requirement under Article 3(1)(b) does not necessitate the CPS in question to be such a gateway for all those users. If Bytedance’s interpretation were correct, it would be sufficient to demonstrate that a CPS is not an important gateway for a specific subset of users, not only in order to call into question the presumption, but also to show that the requirement set out in Article 3(1)(b) of the DMA has not been met.

97. Fourthly, Bytedance argues that in paragraphs 190 to 193 of the judgment under appeal, the General Court committed a legal error in rejecting evidence that TikTok’s focus on video-sharing means that the intensity of its use as an OSN is negligible. Bytedance submits that TikTok’s classification as an OSN does not render any other evidence ineffective in the assessment of whether it is an important gateway.

98. I must admit that I find that this specific critique of the judgment under appeal is not entirely clear. It is uncertain whether Bytedance alleges a misapplication of the important gateway presumption or in reality criticises the assessment of the probative value of the evidence. The title of the relevant section of its appeal suggests the former. In that respect, in paragraphs 190 to 193, the General Court addressed Bytedance’s statement that the Commission should have taken into account that the intensity of use of TikTok ‘as a social network’ is practically non-existent because the principal purpose of that platform is to offer video-sharing services. In paragraph 193, the General Court explained why it considered that the Commission was correct to take into account intensity with regard to the sharing and discovery of videos on TikTok, although the service itself was classified as an OSN and not as a video-sharing platform. In that regard, the General Court relied on the definition of an OSN laid down in point 7 of Article 2 of the DMA, which underscores that one of the essential functions of such a CPS is sharing content, in particular videos.

99. The argument related to a misapplication of law with regard to lock-in effects is therefore unfounded.

100. Bytedance criticises paragraphs 208 to 212 of the judgment under appeal for their consideration of the test applied to assess whether business users depend on that CPS to reach end users, for the purposes of Article 3(1)(b) of the DMA.

101. In that regard, Bytedance submits that, in paragraph 210 of the judgment under appeal, the General Court stated that in order to consider that business users depend on that service, ‘it is not necessary for [the CPS] to be the only channel through which those undertakings reach those users. It is sufficient for [the CPS] to be an important channel for that purpose, which those business users can access only if they have an account on that CPS.’ Bytedance essentially argues that if there are other effective channels to reach the same end users, then that situation of dependency does not arise.

102. In my opinion, the General Court made no error when it stated that, to use the term employed by Bytedance, ‘business dependency’ does not require the service in question to be the sole channel through which business users can reach end users. The DMA does not require extreme business dependency, which could exist, as the Commission points out, only where a CPS holds a monopoly. In my view, it is sufficient, in line with the wording of Article 3(1)(b) of the DMA, that the service in question constitutes an important, but not the only, gateway for business users to reach end users.

103. For those reasons, I take the view that the third complaint of the third part of the first ground of appeal is unfounded.

4. The alleged distortion of evidence and failure to state reasons as regards multi-homing, intensity of use and interoperability

104. The fourth complaint of the third part of the first ground of appeal is composed of multiple arguments alleging distortion of evidence and failure to state reasons.

105. Bytedance argues, first, that, in paragraph 184 of the judgment under appeal, the General Court distorted the evidence that it had produced by stating that ‘all the online social networks to which [Bytedance] refers have high degrees of multi-homing, in so far as more than 65% of the end users of those networks use two other platforms in parallel’. Bytedance submits that the evidence shows no other OSN to which users of Meta services multi-homes over 65% except to another of that undertaking’s platforms.

106. In that regard, for there to be distortion of evidence, the General Court must have manifestly exceeded the limits of a reasonable assessment of the evidence and, moreover, the distortion must be obvious from the documents in the file, without there being any need for the Court to carry out a new assessment of the facts and evidence. (21) In the light of those principles, it must be held that the General Court did not distort the evidence in paragraph 184 of the judgment under appeal.

107. In that paragraph, the General Court indicated that it had reviewed the data on services of ‘all the online social networks’ to which Bytedance itself referred, including YouTube, designated by the Commission as a CPS consisting of a video-sharing platform, and observed that more than 65% of those services’ users use two other services in parallel. Moreover, Bytedance itself not only prompted the inclusion of YouTube in the analysis, but it also argues elsewhere that TikTok’s purpose is to offer video-sharing services. (22) Furthermore, in the data produced by Bytedance, there is no indication as to the importance of the circumstance that Facebook and Instagram are operated by the same undertaking. The remark to that effect appeared only in the appeal. In the light of the figures produced by Bytedance itself, and given that under the presumption-based designation mechanism Bytedance bears the burden of adducing evidence, the General Court did not manifestly distort the evidence to reach its conclusion in paragraph 184 of the judgment under appeal.

108. Secondly, Bytedance claims that paragraph 189 of the judgment under appeal disregards the plain reading of the evidence of usage intensity that it provided. Nonetheless, its argument results from a misreading of that paragraph. The General Court did not state that Bytedance failed to provide data on TikTok’s usage intensity, but that it failed to provide data on the intensity of use of the various OSNs. I would also point out that, in paragraph 199 of the judgment under appeal, the General Court reiterates that statement by agreeing with the Commission’s observation that the figures produced by Bytedance fail to reflect the intensity of use of the various OSNs.

109. Thirdly, Bytedance criticises paragraph 198 of the judgment under appeal for upholding the Commission’s claims that the multi-homing among users of other OSNs results from their longer presence in the internal market. Bytedance argues that there are other non-designated platforms that demonstrate the same degree of multi-homing as TikTok, despite having been operational long before it. I must admit that it is unclear to me whether, in that regard, Bytedance alleges a distortion of evidence or a failure to state reasons. Nevertheless, its argument appears to be that the duration of the service’s presence on the market is not in itself sufficient to explain the phenomena described by the General Court in paragraphs 198 and 200 of the judgment under appeal as ‘the asymmetric nature of multi-homing’, namely the fact that a high proportion of TikTok users also use Facebook and Instagram, whereas, conversely, a lower proportion of users of those services also use TikTok, and that the General Court is incorrect to confirm the decision at issue in that regard.

110. Against that backdrop, before the General Court and now in its appeal, Bytedance essentially argues, as recalled in paragraph 196 of the judgment under appeal, that the asymmetric multi-homing practiced by OSNs’ users implies that TikTok does not benefit from significant network and lock-in effects, which serves to rebut the presumption that it is an important gateway.

111. However, not only is Bytedance incorrect to assert that there is necessarily a link between multi-homing and a lack of lock-in effects, (23) but it also misrepresents the possible alternative explanation of the asymmetrical nature of multi-homing among OSNs’ users. In paragraph 198 of the judgment under appeal, the General Court repeated the point made by the Commission that this asymmetry is likely to be due to the fact that Facebook and Instagram have been in operation for a longer period in the European Union and that they had a large user base in the European Union before TikTok was launched in the internal market. The other non-designated platforms mentioned by Bytedance do not appear to meet the second criterion (having a large user base in the European Union) and it does not appear to argue otherwise.

112. Fourthly, Bytedance alleges that, in paragraphs 199, 244 and 272 to 278 of the judgment under appeal, the General Court distorted the evidence as to intensity of use.

113. First of all, Bytedance argues that, in paragraph 199 of the judgment under appeal, the General Court uncritically adopted the Commission’s unsubstantiated speculation that TikTok has ‘a greater usage intensity among younger users’. Bytedance submits that that speculation proceeds from a misreading of an online article which, in fact, refers to TikTok users increasing in age. Bytedance also adds that, by adopting that speculation as its own, the General Court distorts the evidence by failing to address Bytedance’s data on actual business user intensity and limited social network activity on TikTok.

114. In that regard, as the Commission points out, the article in question not only states that the age of TikTok users is increasing, but it also provides statistics according to which the majority of its users are aged 29 years or younger. (24) Moreover, the article underscores that engagement among TikTok users is ‘incredible’ and that the screen time of its users greatly exceeds that of its competitors. The Commission’s consideration in recital 141 of the decision at issue, confirmed by the General Court in paragraph 199 of the judgment under appeal (‘TikTok enjoyed a higher engagement rate than other social networks since end users, in particular young people, spend more time on TikTok than on other social networks’), must have followed from the combination of those two pieces of information. Admittedly, the article does not appear to state explicitly that the engagement rate of young people is higher than the engagement rate of older users. The article does however demonstrate that the users’ engagement rate with TikTok, including younger people who form the majority of its users, is higher than their engagement with other platforms. On that understanding of paragraph 199 of the judgment under appeal, the finding challenged by Bytedance that TikTok has a greater usage intensity among young users does not proceed from a misreading of that article.

115. Moreover, Bytedance’s allegation that the General Court failed to address the data on actual business user intensity and limited social networking activity on TikTok in paragraph 199 of the judgment under appeal remains unclear. The entire paragraph discusses the engagement of end users, not of business users.

116. Next, Bytedance criticises the General Court for stating, in paragraph 244 of the judgment under appeal, that it produced no evidence of TikTok importance to business users. The argument proceeds from a misreading of the judgment. The General Court found that Bytedance did not put forward any argument challenging specific findings in recitals 151 to 154 of the decision at issue. Furthermore, where a distortion of the evidence is alleged, an applicant must indicate precisely the relevant facts or documents alleged to have been distorted. (25) Bytedance’s argument remains general and abstract and, as such, its allegation of a distortion of evidence cannot succeed.

117. Finally, Bytedance argues that, in paragraphs 272 to 278 of the judgment under appeal, the General Court had no legal basis for rejecting its business user engagement data. Bytedance’s submission is that, contrary to paragraph 273 of that judgment, the decision at issue does not state that Bytedance’s ‘figures are not sufficiently representative for the purposes of determining whether TikTok is an important gateway’. Bytedance argues that recital 149 of the decision at issue examines only the representativeness of the registered businesses data. Bytedance also alleges that the Commission did not engage in a qualitative analysis to examine whether business users considered TikTok an important gateway or to reach any conclusion as to whether non-registered businesses conducted themselves differently to registered businesses. Consequently, in its view, the General Court distorted the evidence in concluding that the activity of registered businesses is not representative of the broader set of business users of TikTok in terms of whether they consider that platform to be an important gateway.

118. It should be borne in mind that the assessment of the facts and evidence and, in particular, the weight to be given to the evidence produced by an applicant come within the scope of the facts of the case and, save where they are distorted, are accordingly not a question of law which is subject to review by the Court of Justice on appeal. (26) Unless an applicant alleges evidential distortion, a submission seeking to challenge the value given to that evidence by the General Court is inadmissible and must be dismissed. (27)

119. In its appeal, Bytedance alleges distortion of evidence but fails to succeed in identifying an error committed by the General Court in that regard. The General Court itself did not draw any conclusions as to whether registered users are more or less likely than other business users to engage with TikTok. Contrary to Bytedance’s allegation, the General Court was also not obliged to engage in such a qualitative analysis, since the burden of adducing evidence that the presumption laid down in Article 3(2)(b) of the DMA does not apply was borne by Bytedance. In that regard, as the Commission points out, paragraph 273 of the judgment under appeal simply reports the finding in the decision at issue that registered business users are a small subset of all TikTok business users and are therefore not representative. Since the data were not representative, it could not have manifestly called into question the presumption laid down in Article 3(2)(b) of the DMA.

120. Fifthly, Bytedance submits that, in paragraphs 206 and 207 of the judgment under appeal, the General Court dismissed evidence of TikTok’s interoperability tools to facilitate multi-homing. Bytedance argues that the General Court stated in paragraph 206 that the evidence is irrelevant because (i) interoperability tools are explained by TikTok’s need to attract users due to its later launch; and (ii) other platforms use similar tools. Bytedance’s submission is that a company that needs interoperability to convince users to try its platform clearly is not a platform that locks users in and the General Court failed to state reasons for reaching the opposite conclusion.

121. In my view, that submission is based, first, on a misreading of the judgment. The General Court neither dismissed the evidence nor found it to be irrelevant; it simply considered that, for the reasons stated by the Commission in recital 142 of the decision at issue (28) and in the light of additional explanation provided by the General Court in paragraph 206 of the judgment under appeal, (29) the interoperability argument was insufficient so as manifestly to call into question the presumption laid down in Article 3(2)(b) of the DMA. As regards, secondly, compliance with the obligation to state reasons, that obligation does not require the General Court to provide an account which follows exhaustively and one by one all the arguments put forward by the parties to the case. That court’s reasoning may therefore be implicit, on condition that it enables the persons concerned to know why it has not upheld their arguments and provides the Court of Justice with sufficient material for it to exercise its power of review. (30) Paragraphs 205 to 207 of the judgment under appeal comply with those requirements, as it explains the reasons that led the General Court not to agree with Bytedance as to the relevance of its interoperability argument.

122. Therefore, in my view, and in the light of the foregoing, the fourth complaint of the third part of the first ground of appeal is unfounded.

5. The alleged misapplication of Article 3(1)(b) of the DMA in relation to relative scale

123. As a preliminary remark, I note that among the elements referred to in recital 23 of the DMA, which the Commission may take into account in its assessment of the arguments seeking to rebut the presumptions laid down inter alia in Article 3(2)(b) of that regulation, is the importance of the CPS having regard to the overall scale of its activities. That recital does not, as the General Court observed in paragraph 222 of the judgment under appeal, specify the parameters on the basis of which the relative importance and scale of the CPS at issue must be measured. However, the General Court observed, in paragraph 223 of the judgment under appeal, that Bytedance had chosen to argue that TikTok is not an important gateway on the ground that it is smaller in size than other well-established platforms.

124. Bytedance argues, in that regard, that, in paragraphs 231 to 235 of the judgment under appeal, the General Court misapplied the legal test for an important gateway and distorted evidence by dismissing as new, and therefore inadmissible, evidence as to TikTok’s size relative to other platforms.

125. More specifically, Bytedance claims, first, that the evidence on TikTok’s relative scale is not new. Bytedance submits that it presented arguments and evidence in that regard with its notification and only subsequently, in response to the Commission’s choice to compare TikTok to two OSNs to assess its ‘relative scale’, did it provide the same ‘repackaged’ data with its action for annulment.

126. As follows from paragraphs 226 to 231 of the judgment under appeal, the arguments and figures that the General Court found to be new concerned (i) the number of TikTok end users compared with the overall scale of activities of OSNs, (ii) the data relating to the overall scale of those activities or (iii) the number of TikTok end users compared with those of Pinterest, LinkedIn, Reddit, DailyMotion or Vimeo.

127. Bytedance fails to show that those specific arguments or figures were presented during the administrative procedure. Admittedly, in its appeal, Bytedance alludes to the elements presented in the course of the administrative procedure and, concerning the fair market value estimations of TikTok and other platforms (Snapchat, Pinterest, Meta, Google, Twitter, Tencent), a timeline showing the launch date of competing services and some general affirmations about alleged TikTok copycats. Nonetheless, those elements by no means correspond, content-wise, to the assertions made in the context of the action for annulment and found by the General Court to be new in paragraphs 226 to 231 of its judgment. Thus, Bytedance has not succeeded in demonstrating that the General Court misapplied the legal test of important gateway or distorted the evidence by dismissing it. In my view, the General Court was fully entitled to find that the evidence was new.

128. Bytedance further reiterates the assertion that the evidence was submitted in response to the Commission’s choice to compare TikTok to just two OSNs. It submits that even if that evidence were new, it would be admissible in the light of case-law (31) ‘in so far as those elements are relevant to the review of the legality of the Commission decision’.

129. However, as the Commission contends and the General Court confirms in paragraph 231 and 243 of the judgment under appeal, it was originally Bytedance’s choice to compare TikTok user numbers to those of Facebook and Instagram, and not to other services or their overall scale of activities. Bytedance simply did not provide any data in the latter regard. It is only in its action for annulment that Bytedance attempted to submit new arguments and evidence and essentially asked the General Court to examine whether, in the light of those arguments and that evidence, the presumption laid down in Article 3(2)(b) of the DMA could be rebutted.

130. Moreover, the interpretation according to which an undertaking could, in order to challenge the legality of a presumption-based designation decision, put forward new arguments and evidence before the General Court is contrary to the scheme and objectives of the DMA.

131. As I have already observed, under the DMA, the burden of adducing evidence to rebut the presumptions is borne by the undertaking concerned. As explained by the General Court in paragraph 233 of the judgment under appeal, Article 3 of the DMA establishes a procedure where all the elements seeking to call into question the presumptions laid down in paragraph 2 thereof are to be presented together with the undertaking’s notification. The General Court also correctly drew attention to Commission Implementing Regulation (EU) 2023/814, (32) Article 2(3) of which requires the undertaking concerned to provide its arguments in an annex to its notification by clearly identifying to which of the three cumulative requirements set out in Article 3(1) of the DMA its arguments relate and, for each argument, to explain why, exceptionally, the CPS concerned does not satisfy that requirement, despite the fact that it meets the corresponding threshold laid down in Article 3(2) of the DMA.

132. Secondly, Bytedance criticises the General Court for upholding, in paragraphs 238 and 243 of the judgment under appeal, the Commission’s approach consisting of comparing TikTok to two other online social networking services, namely Facebook and Instagram, and not to video-sharing platforms. However, TikTok’s classification as an OSN was not contested by Bytedance. More importantly, recital 23 of the DMA, to which Bytedance refers to while discussing the issue of relative scale, underscores the importance of the undertaking’s CPS considering the overall scale of activities of the respective CPS, and not of other services.

133. Thirdly, Bytedance argues that paragraphs 231, 241 and 242 of the judgment under appeal distort the evidence by stating that it did not submit any on the comparative growth of rival platforms during the administrative procedure. In my view, Bytedance’s critique is not compelling. In those paragraphs the General Court considered the number of end users and the growth of user base. By contrast, the evidence to which Bytedance refers in its appeal concerned static figures on its competitors’ user bases in a selected year and some metric of growth relating to views. Thus, as the Commission correctly points out, the data provided by Bytedance did not relate to the growth of user base, but rather to the number of views.

134. In my view, therefore, the fifth complaint of the third part of the first ground of appeal is unfounded.

6. The alleged application of the incorrect test in the ‘important gateway’ analysis

135. The sixth, and final, complaint of the third part of the first ground of appeal consists of two allegations that the General Court failed to apply the correct standard of proof in the judgment under appeal.

136. First, Bytedance argues that, in paragraph 200 of the judgment under appeal, the General Court required it to demonstrate that the asymmetric nature of multi-homing among OSNs users was not due to the economic and historical context. (33) It submits that it should have been required to prove only the absence of significant network and lock-in effects and demonstrate efforts to encourage multi-homing.

137. In my view, Bytedance’s argument results from a misreading of the judgment under appeal. What the General Court alluded to in paragraph 200 of the judgment under appeal was indeed the fact that Bytedance should have demonstrated, regardless of the asymmetric nature of multi-homing, the absence of significant network effects and lock-in effects. The lack of such effects could not be inferred from the asymmetric nature of multi-homing alone, as those effects could be explained by economic and historical context discussed by the Commission in recital 141 of the decision at issue. Such a reading of that paragraph is confirmed by paragraphs 201 to 214, where the General Court explains why Bytedance had failed to demonstrate the lack of significant network and lock-in effects, and not that it had failed to demonstrate that asymmetric multi-homing did not result from the economic and historical context.

138. Secondly, Bytedance submits that, in paragraphs 272 to 279 of the judgment under appeal, the General Court applied the incorrect legal test when confirming the Commission’s finding that the evidence on the engagement of registered business users is insufficiently representative of the broader category of all TikTok business users. Bytedance argues that registered users should be the most engaged and reliant upon the platform to reach end users. The nature of the argument it puts forward is not entirely clear.

139. In that regard, on the one hand, in recital 149 of the decision at issue, the Commission found that the figures on the engagement of registered business users produced by Bytedance were not representative for the purposes of determining whether TikTok is an important gateway. In paragraphs 272 to 279 of the judgment under appeal, the General Court essentially confirmed the Commission’s assessment of that evidence, in response to the argument advanced by Bytedance in the action for annulment. On that reading, Bytedance would appear to be repeating its complaint asserting a distortion of evidence in paragraphs 272 to 278 thereof. (34)

140. On the other hand, Bytedance’s argument could be understood as a critique of the standard of proof applied by the General Court in confirming the Commission’s finding that the arguments and evidence that it presented are not sufficient manifestly to call into question the presumption laid down in Article 3(2)(b) of the DMA. However, in paragraphs 273 to 277, the judgment under appeal explains in considerable detail the reasons that led the General Court to confirm the Commission’s findings as to the insufficiency of those arguments and that evidence in that regard. In essence, the figures relating to registered business users concern only a small portion of all business users, with the result that those figures are not sufficiently representative for the purpose of calling into question the presumption. Consequently, Bytedance’s argument was hardly compelling and revealed an easily identifiable weakness in its capacity to rebut the presumption. In my view, therefore, the General Court’s approach was consistent with the standard of proof required under the DMA.

141. In the light of the foregoing, I consider that the sixth complaint of the third part of the first ground of appeal is unfounded.

D. The second ground of appeal and the Commission’s request for substitution of grounds – right to be heard

142. In this final section, I address Bytedance’s second ground of appeal, which raises the question whether, in a designation procedure under Article 3(4) of the DMA, the Commission is required, by virtue of the right to be heard guaranteed by Article 41(2)(a) of the Charter of Fundamental Rights of the European Union (‘the Charter’), to disclose to the undertaking concerned its provisional assessment of the arguments and evidence submitted by that undertaking before adopting its decision. I consider that it is not.

143. In that regard, by its second ground of appeal, Bytedance argues that paragraphs 340 to 371 of the judgment under appeal infringe Article 41 of the Charter by excusing the breach of its right to be heard. (35) In its response, the Commission formulates a request for a substitution of the relevant grounds of the judgment and argues that the General Court should have concluded that Bytedance’s right to be heard had not been breached.

144. The second ground of appeal and the Commission’s request for a substitution of grounds concern the same aspect of the General Court’s reasoning, namely the scope of the right to be heard in a designation procedure based on the presumptions laid down in Article 3(2) of the DMA. They differ, however, on their starting point. Bytedance challenges the fact that the Commission was excused for infringing the right to be heard by not giving it an opportunity to respond to the Commission’s provisional assessments of the arguments and evidence, while the Commission disputes existence of any breach. I shall therefore start by examining the requirements stemming from the right to be heard in the specific context of the DMA designation procedure (Section 1 below). I shall then apply those requirements to the assessment of the Commission’s request for a substitution of grounds and to Bytedance’s second ground of appeal (Section 2 below).

1. The designation procedure and the right to be heard

(a) Article 34(1) to (3) of the DMA

145. Recital 109 of the DMA (36) recalls that that regulation must be interpreted and applied in accordance with fundamental rights recognised by the Charter. Article 34 of the DMA, entitled ‘Right to be heard and access to the file’, gives specific expression to that requirement by laying down procedural safeguards applicable to decisions adopted by the Commission under the DMA. Article 34(1) thereof provides that before adopting a decision on the basis of the provisions listed therein, the Commission must respect the right of the undertaking concerned to be heard. In the context of a case, in which the quantitative thresholds set out in Article 3(2) of the DMA are met and the Commission, presented with arguments manifestly calling the presumptions into questions, opens a market investigation, Article 34(1) of the DMA, read in conjunction with the second subparagraph of Article 17(3) thereof, requires the Commission, before adopting a designation decision following a market investigation pursuant to Article 17 of that regulation, to communicate its preliminary findings to the undertaking concerned within three months from the opening of the investigation. Article 34(1)(a) of the DMA requires the Commission to give the undertaking concerned the opportunity to be heard on such preliminary findings, while Article 34(2) provides that the undertaking concerned may submit its observations concerning those preliminary findings. Article 34(3) requires the Commission to base its decision only on the preliminary findings on which the undertaking concerned has been able to comment.

146. Article 34(1) of the DMA does not impose the same obligations on the Commission before the adoption of a designation decision made pursuant to Article 3(4) and (5) of the DMA, where no market investigation is opened since the quantitative thresholds of Article 3(2) are met and the arguments challenging the presumptions do not manifestly call them into question. That presumption-based mechanism for a designation decision is simply not referred to in Article 34(1) of the DMA.

147. Furthermore, in the presumption-based designation mechanism under Article 3(4) of the DMA, the Commission is required to designate a gatekeeper without undue delay and at the latest within 45 working days after receiving the complete notification from the undertaking. That time limit applies also when the undertaking challenges the presumptions arising under Article 3(2) of the DMA. By design, there may be insufficient time to communicate any provisional findings to the undertaking, still less to give it an opportunity to comment on them.

148. Thus, in a designation procedure based on the presumptions laid down in Article 3(4) of the DMA, the Commission operates to strict deadlines and its task is limited to assessing the arguments and evidence submitted by the undertaking seeking to rebut those statutory presumptions. The Commission does not conduct an independent investigation into whether the requirements for designation are met unless it decides to open a market investigation. Accordingly, the DMA does not impose, in the context of the presumption-based designation mechanism, the same procedural obligations as those applicable following a market investigation.

149. Similarly, a comparison between the two scenarios described in points 145 and 146 of the present Opinion leads me to conclude that it is not the fact that a procedure culminates in a designation decision that triggers the Commission’s obligations set out in Article 34(1) and (3) of the DMA. Instead, the EU legislature must have considered that those obligations are triggered only when the Commission adopts a designation decision pursuant to either Article 3(4) or Article 3(8) of the DMA, following a market investigation. In the course of such an investigation, the Commission is no longer confined to relying on the arguments and evidence provided by the undertaking concerned and is likely to exercise its investigative powers. (37) The rationale of the Commission’s obligations laid down in Article 34(1) and (3) of the DMA is thus not to give the undertaking the possibility to comment on the institution’s provisional assessment of the arguments and evidence submitted to challenge the statutory presumptions laid down in Article 3(2) thereof, but to comment on the findings that the Commission itself made.

150. Consequently, since the purely presumption-based designation mechanism, such as that applied in the present case, falls outside the scope of Article 34(1) and (3) of the DMA, I shall now consider the right to be heard as it applies in that context.

(b) Article 34(4) of the DMA and Article 41(2)(a) of the Charter

151. Article 34(4) of the DMA provides that the rights of defence of the gatekeeper, undertaking or association of undertakings concerned are to be fully respected in any proceedings. Moreover, as the General Court recalled in paragraph 343 of the judgment under appeal, the rights of defence are fundamental rights forming an integral part of the general principles of law whose observance the Courts of the European Union ensure even where the applicable legislation does not expressly provide for such a procedural requirement. That general principle of EU law is enshrined, in the context of the right to good administration, in Article 41(2)(a) of the Charter and applies where the authorities intend to adopt a measure liable to adversely affect an individual.

152. In the latter regard, I would clarify that Article 41(2)(a) of the Charter applies to the proceedings that culminate in a designation decision adopted on the basis of Article 3(4) of the DMA. The designation as a gatekeeper gives rise to significant legal and operational consequences for the undertaking. A designation decision is therefore liable adversely to affect the interests of the undertaking concerned. Moreover, Article 41(2)(a) of the Charter must be observed not only where the proceedings might lead to penalties, but also where the decision taken may directly and individually affect the undertaking concerned and entail adverse consequences to it. (38)

153. The more delicate question concerns the specific requirements imposed by the right to be heard. In paragraph 344 of the judgment under appeal, the General Court stated that the right to be heard guarantees every person the opportunity to make known his or her views effectively during an administrative procedure and before the adoption of any decision liable to affect his or her interests adversely. In paragraph 349, the General Court recalled that, in accordance with settled case-law, while the right to be heard extends to all the matters of fact and of law on which the decision is based, it does not extend to the administration’s final position.

154. Those general considerations must, however, be viewed in the context of the particular legal framework within which the right to be heard applies. (39)

155. In the specific DMA regulatory framework, both the factual and the legal bases of the Commission’s decision are, to a large extent, known to the undertaking concerned. The relevant facts are provided by the undertaking itself, which is required to submit the information necessary for the Commission and has the burden of adducing evidence capable of rebutting the presumption that it satisfies the requirements under Article 3(1) of the DMA to be designated as a gatekeeper. The applicable legal framework is specifically defined by the DMA itself: the quantitative thresholds triggering the statutory presumptions are laid down in Article 3(2); an undertaking may seek to rebut those presumptions pursuant to Article 3(5), which sets out the standard of proof required to do so; and the Commission must assess the undertaking’s arguments and evidence in rebuttal without having any discretion to open a market investigation to further verify whether the presumed qualitative requirements laid down in Article 3(1) are satisfied.

156. Against that backdrop, first, as already observed by the General Court, (40) the right to be heard does not require that, before taking a final position on the assessment of the evidence submitted by a party, a competent authority must offer that party a further opportunity to comment on that evidence. Secondly, an undertaking which has itself submitted the relevant facts and evidence is ordinarily in a position to address their relevance to the assessment at the time of submission. (41)

157. Thirdly, the Court’s case-law makes clear that, where a competent authority has no discretion and is required to draw a legal consequence from a specific circumstance that is the premiss of its decision, the right to be heard requires that the affected party be given an effective opportunity to express its view on both that circumstance and the applicability of any exception to drawing that legal consequence from it. (42) The right to be heard does not, however, require the authority to disclose its intended decision or its provisional assessment to the affected party. It requires the authority to take into account the observations submitted and to examine carefully and impartially all the relevant aspects of the case, and to give a detailed statement of reasons for its decision. (43)

158. That logic applies, a fortiori, in a presumption-based decision-making process, in which the undertaking itself provides the information demonstrating that the thresholds triggering the presumptions are met and also bears the burden of adducing evidence that, nevertheless, those presumptions do not apply to it. Consequently, in the DMA’s presumption-based designation mechanism, the right to be heard requires that the undertaking have an opportunity to challenge the applicable presumptions. It also requires the Commission to take due account of the arguments and evidence submitted by that undertaking, examining carefully and impartially all the relevant aspects of the individual case and giving a detailed statement of reasons for its decision.

159. Consequently, unless the Commission has used its investigative powers before opening a market investigation, (44) the right to be heard does not require it to give the undertaking an opportunity to comment on its provisional assessment of the arguments and evidence before it adopts its final decision.

160. Applying that distinction may not always be straightforward in practice. The boundary between assessing whether the arguments and evidence submitted by the undertaking are capable of calling a presumption into question, on the one hand, and making its own findings, on the other hand, may be difficult to draw. In the absence of specific guarantees for the protection of the undertaking under the DMA’s presumption-based designation mechanism, (45) the Commission should therefore exercise caution and, to the extent possible, provide the undertaking concerned with information on its preliminary assessments. (46)

161. In the light of those conclusions, I now turn to consider the Commission’s request for a substitution of grounds and Bytedance’s second ground of appeal.

2. Right to be heard in the case at hand

162. The second ground of appeal concerns the alleged breach of Bytedance’s right to be heard. Bytedance argues that, in paragraphs 340 to 371 of the judgment under appeal, the General Court infringed Article 41 of the Charter as it held that that breach of the right to be heard did not warrant setting aside the Commission’s decision in relation to the arguments concerning (a) the existence of an ecosystem (paragraphs 353 to 357 and recital 132 of the decision at issue) and (b) the intensity of use of the service in question (paragraphs 358 to 370 and recital 141 of the decision at issue). Bytedance notes that the General Court found that, while the right to be heard had been infringed, Bytedance failed to demonstrate that the Commission’s decision would have been different had the irregularity not occurred. In that regard, Bytedance considers that the proper test for an unjustifiable breach of the right of defence is not whether the decision at issue could have been different, but whether presenting its views would have enabled it better to defend itself.

163. The Commission formulates a request for the General Court’s grounds to be substituted, contending that it should have concluded instead that the Commission had not infringed Bytedance’s right to be heard. The Commission also observes that it had in fact provided a level of protection going beyond that required by the DMA in that it had communicated to Bytedance its preliminary findings in the letter of 26 July 2023 – to which Bytedance replied by letter of 2 August 2023 – and organised meetings with Bytedance before the notification and on 17 August 2023 after the notification.

164. In view of the Commission’s objection, I shall now examine the findings of the General Court in the light of the interpretation of the right to be heard that I have presented above.

(a) On the existence of an ecosystem

165. As a reminder, Bytedance attempted to challenge TikTok’s presumed gateway position by arguing that, unlike Meta or Alphabet, it lacks a wider ecosystem which it could leverage.

166. In recital 132 of the decision at issue, the Commission considered, however, that Bytedance in fact operated its own ecosystem, consisting of highly popular video-editing services, enterprise software, advertising, news, and healthcare applications. In that respect, in paragraph 353 of the judgment under appeal, the General Court held that the Commission had infringed Bytedance’s right to be heard. The General Court noted that while the absence of an ecosystem and its consequences for the designation as a gatekeeper had been the subject of discussions between the parties, the Commission had not informed the applicant of its view that Bytedance did have an ecosystem.

167. In that regard, first of all, as the General Court observed in paragraph 356 of the judgment under appeal, the Commission’s rejection of the applicant’s arguments alleging the absence of an ecosystem is based on a number of grounds which are founded, with the exception of recital 132 of the decision at issue, on the hypothesis most favourable to it, namely that Bytedance did not have an ecosystem. Those other grounds alone would suffice to justify the Commission’s finding that the applicant’s arguments were not sufficiently substantiated so as manifestly to call into question the presumption laid down in Article 3(2)(b) of the DMA.

168. Next, the Commission’s finding that, contrary to the applicant’s claim, Bytedance operated an ecosystem was based on an assessment of the arguments and evidence submitted by the applicant itself. Admittedly, the Commission could have confined itself to the other grounds mentioned above or to stating that it did not find the arguments and evidence on the absence of an ecosystem compelling. Nonetheless, the Commission decided to respond to Bytedance’s assertion and it simply did not agree; Bytedance had attempted to challenge the presumption laid down in Article 3(2)(b) of the DMA that its service was an important gateway, within the meaning of Article 3(1)(b) of the DMA, by arguing that it does not operate an ecosystem.

169. Against that backdrop, as recalled in point 156 of the present Opinion, the right to be heard does not require that, before taking a final position on the assessment of the evidence submitted by a party, a competent authority must offer that party a further opportunity to comment on that evidence. The Commission was not obliged to disclose to Bytedance its provisional assessment of the evidence and offer it an opportunity to comment on that assessment. If the party does not agree with the final position of the authority, it may challenge that position in an action for annulment before the General Court. That is exactly what Bytedance has done by disputing the Commission’s finding that it had an ecosystem in its action for annulment. (47)

170. Finally, in my view, the Commission also acted with the necessary caution during the administrative procedure. (48) In its letter of 26 July 2023, the Commission indicated that the absence of an ecosystem was not sufficient manifestly to call into question the presumption laid down in Article 3(2)(b) of the DMA.

171. Consequently, there is nothing to suggest that the Commission infringed Bytedance’s right to be heard by not allowing it to comment on the Commission’s finding that it operated an ecosystem, notwithstanding Bytedance’s claim to the contrary.

(b) On the intensity of use of the service in question

172. In paragraph 361 of the judgment under appeal, the General Court observed that the Commission did not dispute the fact that it had not heard Bytedance on the question of the intensity of use of TikTok. Admittedly, in contrast with paragraph 353 of that judgment, in paragraphs 358 to 370 thereof, the General Court does not explicitly state that the Commission infringed Bytedance’s right by not hearing it on that question. However, in paragraph 370, the General Court applied the test used in paragraph 353 to excuse the breach of the right to be heard with regard to the issue of an ecosystem. The Commission itself understands the judgment under appeal as finding that it had infringed Bytedance’s right to be heard in relation to both the ecosystem and the intensity of use issues. I shall likewise proceed on that basis.

173. Bytedance’s submission that that breach cannot be justified must be considered in its wider context. As a reminder, Bytedance challenged the presumed gateway position within the meaning of Article 3(1)(b) of the DMA by arguing, inter alia, that multi-homing by a significant proportion of TikTok business and end users calls into question the presumption laid down in Article 3(2)(b) of the DMA.

174. The Commission considered that prevalent multi-homing was insufficient so as manifestly to call into question that presumption. In recitals 135 to 139 of the decision at issue, the Commission stated that (i) the existence of some degree of multi-homing is not in itself an indication that a CPS is not an important gateway; (ii) multi-homing does not imply equal use of the services in question as the use may be asymmetrical; (iii) even a symmetrical use does not mean that a service is not an important gateway enabling business users to reach their end users, for example, certain end-user demographic groups.

175. In recital 141 of the decision at issue, the Commission then assessed the probative value of the figures produced by Bytedance to show that TikTok users multi-homed in greater proportions than those recorded for Facebook and Instagram. The Commission considered, in that regard, that those figures were of limited evidential value as they could be explained by the mere fact that Facebook and Instagram had been active in the European Union before the arrival of TikTok and that they both had a higher number of end users. The Commission therefore considered that, statistically, a large number of Facebook and Instagram users could be expected also to use TikTok.

176. In the same recital, the Commission also noted that the figures presented by Bytedance do not reflect the intensity of use of those platforms. As the General Court observed in paragraph 362 of the judgment under appeal, the Commission added that it had been reported on numerous occasions that TikTok had a significantly higher engagement rate than other OSNs, that end users spent significantly more time on TikTok than on other OSNs, and that this was the case, in particular, for younger end users. In support of those assertions, the Commission referred to two articles published online, cited in footnotes 152 and 153 to the decision at issue.

177. In that context, the Commission’s references to published articles concerning TikTok’s engagement rate and the time spent by users on the platform cannot be regarded as introducing new facts requiring Bytedance to be heard again before the adoption of the decision. Those references were not used to establish that TikTok enjoyed a particular level of user engagement; they were intended only to illustrate why multi-homing data, considered in isolation, did not necessarily demonstrate the absence of an important gateway position. Recital 141 therefore contains an assessment of the probative value of the material submitted by Bytedance, not an independent finding by the Commission concerning the intensity of use of TikTok.

178. In those circumstances, I therefore consider that the Commission did not infringe Bytedance’s right by not hearing it on the question of intensity of use.

179. In the light of the remarks set out in points 166 and 172 of the present Opinion, the General Court essentially considered that, before the adoption of a designation decision, the Commission is required to communicate to the undertaking concerned its provisional assessment of all arguments and evidence submitted by that undertaking, including its assessment of whether those arguments and that evidence are capable of manifestly calling into question the presumptions laid down in Article 3(2) of the DMA. The Commission must then give the undertaking an opportunity to submit observations on all the grounds on which it intends to rely in its decision.

180. As it transpires from the analysis in point 150 of this Opinion, I take the view that the DMA does not, however, require the Commission, before adopting a presumption-based designation decision, to communicate its provisional assessments to the undertaking concerned and to give it an opportunity to submit observations thereon. In the light of the considerations set out in points 158 to 160 above, nor does any such requirement follow from Article 41(2)(a) of the Charter.

181. Furthermore, I consider that the General Court’s approach to the requirements stemming from that provision could ultimately lead to one of two equally undesirable outcomes.

182. If it were required to communicate to the undertaking concerned each and every element of its assessment of the arguments and evidence submitted by that undertaking, the Commission would find it difficult to comply with the time limit of 45 working days. It would most likely be forced to open a market investigation in each and every case in order to comply with Article 41(2)(a) of the Charter as interpreted by the General Court, and also the tight time limits within which it is required to act.

183. Alternatively, it might be argued that the General Court set the bar unduly high, but paradoxically did so in a manner that provides a justification for any subsequent failure to attain it. If one accepts the logic of the judgment under appeal and considers that the Commission is required to hear the applicant on all the grounds on which it intends to reject the undertaking’s argument and evidence, the Commission could repetitively infringe that right without any effective consequence given the manner in which the General Court excused the alleged breach of the right to be heard, namely by requiring an applicant to demonstrate that, in the absence of the irregularity found, it could not be entirely ruled out that the Commission’s decision would have been different. As the decision at issue illustrates, the Commission does not appear, in its decision-making practice, to rely on a single ground to explain why it considers the undertaking’s arguments and evidence insufficient to call into question the presumptions operating against it. Consequently, if the General Court’s reasoning were accepted, the undertaking concerned would be unable to demonstrate that the designation decision would have been different; a failure to hear the undertaking on one of the many grounds of its decision would not affect the outcome of the procedure. It is inconceivable, however, that the EU legislature could have intended the DMA to establish a mechanism that tolerates repeated breaches of the right to be heard without affording the party concerned any effective means of challenging those breaches.

184. Overall, it follows that the General Court erred in law in finding that the Commission was required to give Bytedance the opportunity to submit observations on its findings in recitals 132 and 141 of the decision at issue. That said, it is also apparent that, in paragraphs 353 and 370 of the judgment under appeal, the General Court was entitled to conclude that the arguments alleging a breach by the Commission of Bytedance’s right to be heard in respect of those recitals had to be rejected.

185. On the basis of the foregoing considerations, I conclude that the second ground of appeal alleging a breach of the right to be heard should be rejected as unfounded and that the appeal should be dismissed in that regard, on the legal grounds set out in points 167 to 170 and 177 of the present Opinion, which should be substituted for those relied on by the General Court in paragraphs 353 and 370 of the judgment under appeal.

IV. Conclusion

186. In the light of the foregoing considerations, I propose that the Court should dismiss the first and third parts of the first ground of appeal. I further propose that the Court should dismiss the second ground of appeal, on the basis set out in points 167 to 170 and 177 of the present Opinion, which should be substituted for those adopted by the General Court in paragraphs 353 and 370 of the judgment of 17 July 2024, Bytedance v Commission (T‑1077/23, EU:T:2024:478).

1 Original language: English.

2 Judgment of 17 July 2024 (T‑1077/23, ‘the judgment under appeal’, EU:T:2024:478).

3 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; ‘the DMA’).

4 See also points 19 and 20 of the present Opinion.

5 See recital 16 of the DMA.

6 While the Commission’s decision-making practice cannot be binding on the Court as regards the interpretation of Article 3(5) of the DMA, it may nevertheless constitute a relevant contextual element when assessing how the mechanism established by that provision functions.

7 For example, Decision of 5 September 2023 relating to a decision pursuant to Article 3 of [the DMA] (Cases DMA. 100017 – Microsoft – Online Social Networking Services; DMA.100023 – MICROSOFT – Number-Independent Interpersonal Communications Services; DMA.100026 – Microsoft – Operating System) (C/2023/551), recital 134; and Decision of 13 May 2024 relating to a decision pursuant to Article 3 of [the DMA] (Case DMA.100042 – ByteDance – Online advertising services) (C/2024/6122), recitals 48 and 60.

8 Decision of 5 February 2026 relating to a letter concerning Apple’s notifications under Article 3(3) of [the DMA] (Cases DMA.100014 – Apple – Online intermediation services – Verticals; DMA.100045 – Apple – Online advertising services) (C/2026/2289), recitals 65 and 107.

9 See point 28 of the present Opinion.

10 See point 34 of the present Opinion.

11 A similar view has already been expressed in the literature: ‘the legal standard for the [Commission] to decide whether to open a market investigation is very high’, in Ribera Martínez, A., ‘The requisite legal standard of the Digital Markets Act’s designation process’, Journal of Competition Law & Economics, 2024, Vol. 20(4), pp. 265 to 291, at p. 286.

12 See point 36 of the present Opinion.

13 See recital 13 of the DMA: ‘Weak contestability and unfair practices in the digital sector are more frequent and pronounced for certain digital services than for others. This is the case in particular for widespread and commonly used digital services that mostly directly intermediate between business users and end users and where features such as extreme scale economies, very strong network effects, an ability to connect many business users with many end users through the multisidedness of these services, lock-in effects, a lack of multi-homing or vertical integration are the most prevalent.’ Moreover, recital 23 of the DMA explains that, in its assessment of the evidence and arguments produced to challenge the presumptions, the Commission should take into account, inter alia, how much the actual business user and end user numbers exceed the thresholds by.

14 See point 64 of the present Opinion.

15 See also point 65 of the present Opinion.

16 In particular, in recital 136 of the decision at issue, the Commission referred to recital 13 of the DMA and rejected Bytedance’s claim that this regulation suggests that extensive multi-homing demonstrates a manifest absence of lock in. In recital 138, the Commission stated that multi-homing does not exclude switching costs if a platform user focuses the activity on one of the platforms.

17 See paragraph 163 of the judgment under appeal.

18 Recital 141 of the decision at issue states, in particular, that ‘it has been repeatedly reported that TikTok has a significantly higher engagement rate than other online social networking services, with end users spending significantly more time on TikTok as compared to other online social networking services [in particular,] younger end users.’ The argument also appears in recital 139: ‘even if some end or business users were similarly active[, it] does not mean that a specific online social networking service is not, individually, a gateway for business users to reach end users (e.g., certain end user demographics).’

19 See, by analogy, judgment of 25 March 2021, Generics (UK) v Commission (C‑588/16 P, not published, EU:C:2021:242, paragraphs 127 to 129).

20 See points 86 and 93 of the action for annulment.

21 Judgment of 29 October 2015, Commission v ANKO (C‑78/14 P, EU:C:2015:732, paragraphs 54 and 55).

22 See point 97 of the present Opinion.

23 See point 89 of the present Opinion.

24 In that regard, the article reads as follows: ‘The percentage of U.S.-based TikTok users by age: 10-19 – 32.5%, 20-29 – 29.5%, 30-39 – 16.4%, 40-49 – 13.9%, 50+ – 7.1%. This means that the TikTok userbase is aging up (so get on it now!). All data via Comscore.’ The Commission reads the data as demonstrating that 62% of users are aged 29 years or younger. I would add that the article also states elsewhere that 60% of TikTok users are ‘Gen Zers’ and that 67% of teens use TikTok.

25 See, to that effect, judgment of 20 October 2011, PepsiCo v Grupo Promer Mon Graphic (C‑281/10 P, EU:C:2011:679, paragraph 78).

26 Judgment of 26 September 2013, Centrotherm Systemtechnik v OHIM and centrotherm Clean Solutions (C‑610/11 P, EU:C:2013:593, paragraph 44).

27 See judgment of 22 November 2012, E.ON Energie v Commission (C‑89/11 P, EU:C:2012:738, paragraph 101).

28 See paragraph 205 of the judgment under appeal.

29 See point 120 of the present Opinion.

30 Judgment of 10 April 2014, Areva and Others v Commission (C‑247/11 P and C‑253/11 P, EU:C:2014:257, paragraph 55).

31 In that regard, Bytedance refers to judgments of 21 January 2016, Galp Energía España and Others v Commission (C‑603/13 P, EU:C:2016:38, paragraph 72); of 9 June 2016, PROAS v Commission (C‑616/13 P, EU:C:2016:415, paragraph 43); and of 27 June 2024, Servier and Others v Commission (C‑201/19 P, EU:C:2024:552, paragraph 124).

32 Implementing Regulation of 14 April 2023 on detailed arrangements for the conduct of certain proceedings by the Commission pursuant to [the DMA] (OJ 2023 L 102, p. 6).

33 The reference to the economic and historical context concerns the fact that other OSNs (namely Facebook and Instagram) were operational in the internal market long before TikTok, and those services had a larger user base in the European Union before TikTok was launched in that market. See points 109 to 111 of the present Opinion.

34 See point 117 of the present Opinion.

35 In that respect, while Bytedance’s appeal and the judgment under appeal occasionally make reference to the rights of the defence, by referring to the infringement of Article 41 of the Charter, Bytedance essentially submits that it was not offered an opportunity to present its views on some issues and thus its right to be heard enshrined in Article 41(2)(a) of the Charter has been infringed. Consequently, I shall analyse the second ground of appeal from the perspective of the latter right.

36 Recital 109 of the DMA clarifies that ‘this Regulation respects the fundamental rights and observes the principles recognised by the Charter of Fundamental Rights of the European Union, in particular Articles 16, 47 and 50 thereof. Accordingly, the interpretation and application of this Regulation should respect those rights and principles’.

37 Arguably, under Article 16(2) of the DMA, the Commission may exercise its investigative powers before opening a market investigation. However, I am not convinced that the Commission should resort to those measures in the assessment of the arguments and evidence submitted with the notification pursuant to Article 3(3) and (5) of the DMA. In principle, if the Commission establishes the need to carry out further verifications via the exercise of its investigative powers, it should open a market investigation in accordance with Article 17(3) of that regulation.

38 See, to that effect, in the context of anti-dumping measures, judgment of 27 June 1991, Al-Jubail Fertilizer v Council (C‑49/88, EU:C:1991:276, paragraph 15).

39 See, to that effect, judgment of 11 December 2014, Boudjlida (C‑249/13, EU:C:2014:2431, paragraph 35).

40 See point 153 of the present Opinion.

41 See, by way of analogy, judgment of 4 March 2020, Tulliallan Burlington v EUIPO (C‑155/18 P to C‑158/18 P, EU:C:2020:151, paragraph 94).

42 See, to that effect, judgment of 11 December 2014, Boudjlida (C‑249/13, EU:C:2014:2431, paragraphs 47 and 55).

43 See, to that effect, judgment of 11 December 2014, Boudjlida (C‑249/13, EU:C:2014:2431, paragraph 38).

44 See footnote 37 to the present Opinion.

45 See points 146 and 150 of the present Opinion.

46 See, to that effect, in the context of anti-dumping measures, judgment of 27 June 1991, Al-Jubail Fertilizer v Council (C‑49/88, EU:C:1991:276, paragraph 16), where the Court held that ‘it should be added that, with regard to the right to a fair hearing, any action taken by the Community institutions must be all the more scrupulous in view of the fact that, as they stand at present, the rules in question do not provide all the procedural guarantees for the protection of the individual which may exist in certain national legal systems.’

47 See paragraphs 136 to 145 of the judgment under appeal.

48 Concerning the necessary caution in the Commission’s actions, see point 160 of the present Opinion.

Opinion of Advocate General Biondi delivered on 15 October 2026.###

Full text

Continue your research in ChatGPT or Claude