ECLI:EU:C:2025:1000

Full text

Provisional text

OPINION OF ADVOCATE GENERAL

RANTOS

delivered on 18 December 2025 (1)

Case C‑145/24 P

BdM Banca SpA, formerly Banca Popolare di Bari SpA

European Commission

( Appeal – State aid – Aid granted by the Italian authorities to Banca Tercas – Intervention by the Fondo interbancario di tutela dei depositi (Interbank Deposit Protection Fund, Italy) (FITD) to cover Banca Tercas’ losses and to support its sale to Banca Popolare di Bari – Commission Decision (EU) 2016/1208 declaring the measure to be State aid incompatible with the internal market – Decision annulled by the judgment of 19 March 2019, Italy and Others v Commission (T‑98/16, T‑196/16 and T‑198/16, EU:T:2019:167) – Judgment upheld on appeal by the judgment of 2 March 2021, Commission v Italy and Others (C‑425/19 P, EU:C:2021:154) – Action for damages – Article 340 TFEU, second paragraph – Limitation period – Concept of a ‘sufficiently serious breach’ of EU law – Causal link )

Introduction

1. By its appeal, BdM Banca SpA, formerly Banca Popolare di Bari SpA (‘the appellant’) requests that the judgment of the General Court of the European Union of 20 December 2023, Banca Popolare di Bari v Commission (T‑415/21, EU:T:2023:833) (‘the judgment under appeal’) be set aside, by which the General Court dismissed its action seeking compensation for the damage caused by Commission Decision (EU) 2016/1208 of 23 December 2015 on State aid SA.39451 (2015/C) (ex 2015/NN) granted by Italy to the bank Tercas (‘the Tercas decision’). (2)

2. This appeal invites the Court to examine common aspects relating to actions for non-contractual liability of the European Union such as, first of all, the starting point of the limitation period provided for in Article 46 of the Statute of the Court of Justice of the European Union, in particular as regards the distinction between ‘immediate’ and ‘continuous’ damage; next, the possible classification of a breach of Article 107(1) TFEU as ‘sufficiently serious’ for the purposes of Article 340 TFEU, second paragraph; and, lastly, the assessment of the causal link between the conduct complained of (here: conduct by the Commission) and the damage allegedly suffered by the appellant, specifically where there are possible concomitant causes of that damage.

Background to the dispute and the judgment under appeal

3. By the Tercas decision, the Commission took the view that the intervention by the Fondo interbancario di tutela dei depositi (Interbank Deposit Protection Fund, Italy) (‘the FITD’), authorised on the 7 July 2014 by the Banca d’Italia (Bank of Italy), for the benefit of Banca Tercas (‘Tercas’), (3) the entirety of the capital of which had been held by the appellant since 1 October 2014, constituted State aid incompatible with the internal market within the meaning of Article 107(1) TFEU that had to be recovered from its beneficiary. (4) On 4 February 2016, the FITD carried out a ‘voluntary intervention’ for the benefit of Tercas and, on 14 July 2016, the appellant acquired Tercas.

4. By judgment of the General Court of 19 March 2019, Italy and Others v Commission, (5) upheld on appeal by the judgment of the Court of Justice of 2 March 2021, Commission v Italy and Others (6) (together, ‘the Tercas judgments’), the Tercas decision was annulled. Following those judgments, the appellant applied to the Commission for compensation for the damage it had allegedly suffered as a result of the adoption of the Tercas decision. The Commission rejected that application.

5. By the judgment under appeal, the General Court dismissed the action brought by the appellant seeking compensation for that damage. After finding that the conditions relating to the unlawfulness of the conduct and to the existence of a causal link were not satisfied, (7) the General Court concluded that that action had to be dismissed, without it being necessary to consider the condition that actual damage was suffered. (8)

Procedure before the Court of Justice and forms of order sought

6. On 26 February 2024, the appellant brought an appeal against the judgment under appeal. It claims that the Court of Justice should, first, set aside the judgment under appeal and, second, primarily, give final judgment in the case, order the European Union, represented by the Commission, to pay it EUR 203.3 million in compensation for the material damage and an appropriate amount of compensation for the non-material damage, and order the Commission to pay the costs; or, in the alternative, refer the case back to the General Court.

7. The Commission contends that the Court should dismiss the appeal as inadmissible or ineffective and, in any event, as being unfounded in its entirety or, in the alternative, dismiss the claim for compensation as wholly unfounded, and order the appellant to pay the costs.

8. The parties presented oral argument and replied to questions put by the Court at the hearing held on 30 September 2025.

Analysis

9. The appeal is divided into three grounds concerning, first, the limitation period of the action in matters arising from non-contractual liability, within the meaning of Article 46 of the Statute of the Court of Justice of the European Union, in respect of one of the heads of damage; second, the classification of the alleged breach of Article 107(1) TFEU as ‘sufficiently serious’ for the purposes of Article 340 TFEU, second paragraph; and, third, the existence of a causal link between the conduct alleged against the Commission and the damage allegedly suffered by the appellant within the meaning of the latter provision.

The first ground of appeal

10. The first ground alleges infringement of Article 46 of the Statute of the Court of Justice of the European Union, in that the General Court held that the action was inadmissible in so far as it concerned damage associated with the reduction in staff numbers. The appellant criticises the General Court for having taken the view that that damage had materialised on the adoption of the incentive scheme that was put in place on 30 December 2015 (‘the incentive scheme’) (9) and was instantaneous in nature, (10) whereas, in fact, that damage was not the result of that scheme but rather of the subsequent transactions which led to the actual reduction in staff and it was therefore continuous in nature. (11)

11. Under Article 46 of the Statute of the Court of Justice of the European Union, which applies to the procedure before the General Court in accordance with the first paragraph of Article 53 of that statute, proceedings against the Union in matters arising from non-contractual liability are to be barred after a period of five years from the occurrence of the event giving rise thereto. (12) According to the case-law of the Court, that period begins to run once the requirements governing the obligation to provide compensation for damage are satisfied and, in particular, once the damage to be made good has materialised. (13) Thus, the period does not begin to run on the date on which the harmful event occurred, but from when the loss actually materialises in relation to the persons concerned. (14)

12. In paragraph 61 of the judgment under appeal, the General Court simply observed that the damage associated with the reduction in staff numbers resulted from the incentive scheme, which was put in place on 30 December 2015, and that therefore the limitation period had expired when the appellant made its application to the Commission, without providing any explicit explanations as to the nature of the damage in question.

13. In that regard, I note that, as the incentive scheme was binding in nature, the alleged damage materialised on the date of that scheme, (15) with the fact that it was implemented at a later point in time being irrelevant. (16) ‘Continuous damage’ should be understood to mean damage that does not occur in the course of a single event but which rather appears to develop progressively, the scale of which increases in proportion to the time which has elapsed. (17) In other words, the cause of the damage is not a single event but rather a situation which persists, whereas, in the present case, the alleged damage remains fixed from as early as the adoption of the incentive scheme.

14. It is true that, on the date when the incentive scheme was adopted, the appellant was unable to quantify the exact amount of the damage, which could vary according to the success or the failure of that scheme. However, on that same date, it was able to assess the maximum amount of the alleged damage and to allocate the resources needed to address that eventuality. Furthermore, Article 340 TFEU, second paragraph, does not prevent the Court from being asked to declare the European Union liable in principle for imminent damage foreseeable with sufficient certainty even if the damage cannot yet be precisely assessed. (18)

15. Moreover, even taking the view, as the appellant does, that the alleged damage began to materialise ‘from the first disbursement’ or that its effects were felt gradually as employees signed up to the incentive scheme, those circumstances do not mean, however, that that damage can be characterised as ‘continuous’. In my view, in such a situation, first, there is no legal basis for distinguishing between the first and subsequent disbursements (19) and, second, that damage consists rather in a serious of losses linked to the successive agreements by employees to sign up to the incentive scheme, with each loss being subject to its own limitation period.

16. In that regard, however, as the Commission observes in its pleadings, it is sufficient to state that the appellant defined its claim for compensation by referring to the funds earmarked in the balance sheet dated 31 December 2015 and has not produced any evidence relating to the various dates on which employees signed up to the incentive scheme offered by the appellant.

17. In the light of the foregoing, I consider that the first ground of appeal should be dismissed.

The second ground of appeal

18. The second ground alleges infringement of the second paragraph of Article 340 TFEU, in so far as the General Court found that the Commission’s unlawful conduct did not constitute a ‘sufficiently serious breach’ of EU law and it did not correctly apply the criterion of a serious and manifest breach of that law. According to the appellant, first, the Commission had no discretion to evaluate whether the FITD’s intervention for the benefit of Tercas constituted State aid within the meaning of Article 107(1) TFEU or, in any event, had limited discretion and, secondly, the General Court should not have taken into consideration the legal and factual complexity of the case forming the subject of the Tercas decision or, in any event, should have ruled out such complexity in the present case.

19. In the judgment under appeal, the General Court held that, despite the error of assessment attributable to the Commission, (20) the Commission had not committed a sufficiently serious breach of Article 107(1) TFEU, because its unlawful conduct was not alien to the normal, careful and diligent conduct of an institution entrusted with the task of overseeing the application of competition rules. (21)

20. As a preliminary point, I recall that, according to settled case-law of the Court, a sufficiently serious breach of a rule of law that is intended to confer rights on individuals, which is one of the conditions that must be satisfied for the European Union to incur non-contractual liability, is established where that breach implies that the institution concerned manifestly and gravely disregarded the limits set on its discretion. (22) Thus, the identification of such a breach presupposes that an irregularity is found that would not have been committed in similar circumstances by an administrative authority exercising ordinary care and diligence. (23)

21. In addition, in order to determine whether a breach of a rule of EU law is to be regarded as sufficiently serious, reference must be made to the field, circumstances and context in which the institution acts. The factors to be taken into consideration in that connection are, inter alia, the degree of clarity and precision of the rule breached and the measure of discretion left by that rule to the EU authorities, the complexity of the situation to be regulated, difficulties in the application or interpretation of the legislation, and whether any error of law was excusable or inexcusable. (24)

22. In that regard, the measure of discretion left by the rule of law breached to the EU authority is only one of the factors to take into consideration in order to determine whether that authority committed a sufficiently serious breach of that rule. While that is a relevant factor, which must be assessed in every case, the fact that the provision breached does not leave any discretion does not necessarily mean that that breach is sufficiently serious. Depending on the circumstances of each case, other factors may be taken into account, having regard to the context in which the infringement was found to have been committed. Accordingly, a breach of a rule of law that leaves no discretion to the authority concerned may not appear, in the light of the circumstances, to be manifest and therefore sufficiently serious, in particular if it results from an error of law that may be excused by having regard to difficulties interpreting the legislation containing that rule. (25)

23. Consequently, the liability test applied by the Court includes, first, a subjective element concerning the discretion left to the institution and, second, objective elements relating to the context. (26) While simply infringing EU law may, in certain circumstances, establish the existence of a sufficiently serious breach when the rule breached leaves the EU authority which committed that infringement only a reduced or even no discretion, (27) such a finding may follow only from all the circumstances surrounding that infringement, where examination of those circumstances does not disclose any other relevant factor that may make it possible to find that the breach of that limit on discretion was not manifest and serious. (28)

24. In the present case, it was held in the Tercas judgments that the unlawfulness of the Tercas decision stemmed, in essence, from two conceptual errors linked, first, to confusion between the condition relating to the imputability of aid and that relating to State resources (29) and, secondly, to the assessment of the condition of the imputability to the State of the aid, granted in the present case by a private entity, using the same test applied when aid is granted by a public undertaking. (30)

25. With regard, in the first place, to the confusion between the condition relating to the imputability of aid and that relating to State resources, I note that, in the Tercas decision, the Commission did not draw a distinction between those two conditions because the indicators connected with the conditions were, in its view, identical.

26. In that regard, I note that it is not unusual for the Commission, like the Courts of the European Union, to assess certain criteria for State aid jointly, where the points of fact examined are common to those conditions. (31) That is particularly the case with the conditions relating to the advantage for the recipient and the selectivity of the measure (for example, as far as concerns tax aid), (32) the conditions for the distortion of competition and effects on trade between Member States (33) and, as is relevant here, the conditions for the imputability of a measure to the State and the use of State resources. (34)

27. As regards, in the second place, the assessment of the condition of the imputability of aid to the State, it is true that, as is apparent from the Tercas judgments, (35) the Commission could not, in the present case, rely on the unlikelihood of an absence of actual influence and control by the public authorities over the FITD, but rather was required to establish that the aid measure under consideration had been adopted under the actual influence or control of the public authorities. According to the Courts of the European Union, the Commission essentially applied an incorrect test.

28. That said, I consider that, in the circumstances of the present case, the Commission was called upon to examine a complex set of indicators in order to establish the degree of involvement of the public authorities in granting the measures which had been provided by an entity, such as the FITD, the ‘private’ nature of which was not apparent from the outset, in particular in the light of its hybrid nature and its ties with the Bank of Italy. (36) As the General Court correctly acknowledged in paragraph 120 of the judgment under appeal, the Commission should have applied the concept of ‘State aid’ within the meaning of Article 107(1) TFEU in the particularly complex legal and factual context, in which the aid measures were granted by a private entity, assessing accordingly the circumstances and information which supported the inference that the measure could be imputed to the State, the factual and legal context of the national measures addressed in the Tercas decision, the involvement of State representatives at the various stages of the intervention and the public mandate with which the FITD had been entrusted.

29. Furthermore, although the concept of ‘State aid’ for the purposes of Article 107(1) TFEU is an objective concept, the interpretation of its criteria may, even if it does not require complex factual or economic analyses, (37) raise complex legal questions, as is the case here. (38) Even without any discretion, the examination of the facts can, by its very nature, entail difficulties in the interpretation and, in particular, the application of the underlying legal concepts, and lead to conceptual errors such as those which the Commission is alleged to have made in the present case. (39)

30. I am therefore of the view that, even though the Commission had a reduced or even no discretion in the present case, the complexity of the situation which it examined remains relevant for determining whether that institution committed a sufficiently serious breach of EU law. (40) Accordingly, the relevant criterion, when assessing the existence of a ‘sufficiently serious breach’, remains, in any case, that of the conduct of a diligent administration.

31. It follows that the errors committed by the Commission in the present case can be characterised as ‘mere errors of assessment’, which cannot be sufficient, as such, to find a manifest and serious breach of EU law – or, in other words, an irregularity that would not have been committed in similar circumstances by an administrative authority exercising ordinary care and diligence – capable of rendering the European Union liable. (41)

32. In the light of the foregoing, I propose that the second ground of appeal be dismissed.

The third ground of appeal

33. The third ground of appeal alleges infringement of the Article 340 TFEU, second paragraph, and of Article 91(e) and Article 96 of the Rules of Procedure of the General Court, in that the General Court found there to be no causal link between the Commission’s unlawful conduct and the alleged damage, that is to say, a loss of confidence in the appellant on the part of its customers. (42)

34. In accordance with the case-law of the Court, the condition relating to the existence of a causal link concerns, in essence, a sufficiently direct causal nexus between the conduct of the EU institutions and the damage, the burden of proof of which rests on the appellant, so that the conduct complained of must be the determining cause of the damage. (43)

35. In the present case, the appellant alleges that the General Court, in the first place, found – in order to rule out the existence of a causal link between the Commission’s conduct and the heads of damage relating to the loss of customers – that the appellant had neither distinguished its own customers from those of Tercas nor clarified the scale of each of those groups.

36. However, contrary to what the appellant claims, the General Court did not simply find that the appellant had drawn no distinction between its own customers and those of Tercas, rather it held that nor had the appellant put forward specific arguments as to whether Tercas’ loss of customers and direct deposits could be attributed to the Tercas decision, (44) which the appellant has not contested.

37. In any event, in the judgment under appeal, the General Court held that the decisions of the appellant’s customers, which caused the alleged damage, had been taken on the basis of assessments and evaluations they made having regard to their own financial interests, in view of the following circumstances: first of all, the appellant’s customers had been placed under no legal obligations by the Tercas decision; next, that decision did not contain anything suggesting that the appellant was unable to adopt alternative measures for the benefit of Tercas or seeking to reduce the appellant’s credibility or diminish its customers’ confidence in it; lastly, as soon as the Tercas decision was announced, both the Italian Government and the appellant had indicated that voluntary intervention measures were already ready and that, consequently, there would be no adverse effects resulting from the Tercas decision. (45)

38. In that regard, the appellant merely criticises the General Court’s conclusions, without demonstrating that they are based on an error in law or on a distortion of the facts or the evidence.

39. In the second place, the appellant alleges that the General Court wrongly made reference to the judgment in Fondazione Cassa di Risparmio di Pesaro and Others v Commission. (46)

40. In that regard, it is sufficient to note that the General Court referred to that judgment simply to reject the appellant’s argument that the judgment confirmed the appellant’s position and concluded that the judgment was not sufficient to establish the existence of a causal link. Therefore, the criticisms made by the appellant cannot call into question the assessment of the General Court, which is not based on that judgment.

41. In the third place, the appellant criticises the General Court for having found that different factors contributed to causing the alleged damage.

42. In that regard, I recall that, according to the case-law of the Court, it is necessary that the alleged damage was actually caused by the conduct alleged against the institutions. Even in the case of a possible contribution by the EU institutions to the damage for which compensation is sought, that contribution might be too remote because of other factors, such as the responsibility resting on others, possibly the appellants. (47) Furthermore, although the question as to whether there is a causal link between the wrongful act and the damage is a question of law and thus is subject to review by the Court of Justice in the context of an appeal, (48) the General Court’s factual assessment of the causal link falls, however, outside the jurisdiction of the Court of Justice in hearing an appeal, except where the facts or evidence have been distorted. (49) Such distortion can be found to exist only where it is obvious from the documents in the Court’s file, without there being any need to carry out a new assessment of the facts and the evidence, that the assessment of the existing evidence is clearly incorrect. (50)

43. In the present case, the appellant contests four of the five factors taken into account by the General Court, in paragraphs 144 to 149 of the judgment under appeal, to reject the existence of a causal link between the Commission’s conduct and the alleged damage: first, the existence of losses recorded in the 2015 balance sheet; second, the lack of information provided to investors, which resulted in administrative penalties being applied and criminal investigations initiated; third, the fact that Tercas was already in difficulty during 2013; and, fourth, the fact that the expenditure generated by the measures to mitigate the allegedly adverse effects of the Tercas decision could not be attributed to that decision. (51)

44. As a preliminary point, I consider that, contrary to the appellant’s arguments, the effect of potential concomitant causes is not necessarily limited to the quantification of the damages but may, possibly, affect the very existence of a causal link between the conduct complained of and the alleged damage.

45. First, with regard to the losses recorded in the 2015 balance sheet, the appellant claims that, according to its own experts, those losses could have contributed at the very most to a proportion of the losses of customers, which those experts estimate at 50%. (52) However, it fails to demonstrate that the judgment under appeal is vitiated by a distortion of facts or evidence, and nor does it contest the other contextual elements taken into account by the General Court in that regard: first of all, the fact that, since the acquisition of Tercas in the course of 2014, the appellant’s financial indicators had not stopped deteriorating and that the reduction of the unit value of the shares held by it, decided upon in response to and at the same time as the approval of the appellant’s 2015 annual report (specifically, on 24 April 2016), had provoked discontent among its customers; (53) next, the finding that, prior to the adoption of the Tercas decision, specifically between the months of December 2014 and December 2015, the appellant had already lost 4.9% of its direct deposits; (54) lastly, the fact that the reform of the cooperative banks, which the appellant itself identified as one of the reasons why it had to review its 2015-2019 business plan and replace it with a new 2016-2020 plan, also contributed to eroding the confidence of its customers. (55)

46. Secondly, with regard to the fact that, according to the appellant, the penalties imposed on its directors were adopted after the damage allegedly occurred, it must be stated that, in the judgment under appeal, the General Court found that it was clear from the decisions of the Commissione Nazionale per le Società e la Borsa (National Companies and Stock Exchange Commission, Italy) (Consob) that, in the context of the capital increases that took place between November 2014 and June 2015, the appellant had failed to inform investors of the method used and had thus set the share price at a higher level than that set by the expert appointed to determine the share price, with the result that directors of the appellant had been made the subject of administrative penalties and criminal investigations, beginning in 2017. (56) Therefore, according to the General Court’s assessment, those penalties, adopted after the period in which the damage allegedly materialised, simply confirmed the appellant’s earlier conduct, which was likely the root cause of the loss of confidence on the part of its customers.

47. Thirdly, as regards Tercas’ earlier insolvency, (57) the appellant claims that that insolvency cannot have represented a determining cause of the damage in question, since it had been a known fact for a long time and that, on account of that state of insolvency, the process of integrating Tercas within the appellant had been preceded, since 2014, not only by the FITD’s full coverage of Tercas’ financial deficit, but also by additional interventions to consolidate the equity of both Tercas and the appellant. However, it should be pointed out that those arguments do not demonstrate that Tercas’ financial situation was not one of insolvency, nor that the General Court distorted facts or evidence in that regard.

48. Fourthly, as for the expenditure generated by the measure to mitigate the allegedly adverse effects of the Tercas decision, (58) the appellant has failed to demonstrate that, if that decision had not been adopted, there would not have been a decrease in deposits or losses of customers, or that the remedial measures had actually been adopted in response to losses specifically linked to the decision and not to other facts.

49. In the fourth place, the appellant criticises the General Court for having found the technical reports produced by the appellant to be unreliable, namely one report from an audit firm and the other from a university professor, without having commissioned an expert’s report within the meaning of Article 91(e) and Article 96 of the Rules of Procedure of the General Court.

50. However, I note that, in the judgment under appeal, the General Court did not simply call into question the reliability of those reports. On the contrary, it took the view, first, that the technical report of the audit firm did not prove that the Tercas decision had been the direct and decisive cause of that damage; secondly, that the reliability of the technical report by the university professor was weakened by the fact that that professor himself conceded that the applicant had suffered a significant loss in the course of 2015 and that, as early as the end of 2016, the Tercas decision was merely one factor amongst others which caused the alleged damage; and, thirdly, that those two reports, which were, moreover, based on documents produced by the appellant, did not contain any legal analysis of the causal link between the Tercas decision and the alleged damage. (59)

51. In any event, it appears to me that, again, the appellant simply challenges the General Court’s assessment of the points of fact and proposes a new assessment, without however demonstrating, or even pleading, that the facts or evidence have been distorted.

52. In addition, with regard to the argument according to which the General Court dismissed the technical reports in question without commissioning an expert’s report, it should be observed that assessment of the appropriateness of adopting a measure of organisation of procedure or a measure of inquiry is a matter falling exclusively within the remit of the Courts of the European Union, which are not required to adopt such a measure where they consider that the evidence produced by the parties is insufficient to assess the existence of the causal link.

53. In the light of the foregoing, I propose that the third ground of appeal be dismissed.

Conclusion

54. In the light of all of the foregoing considerations, I propose that the Court dismiss the appeal.

1 Original language: French.

2 OJ 2016 L 203, p. 1.

3 The action taken by the FITD consisted in three aid measures: first, a EUR 265 million contribution intended to cover Tercas’ negative equity, second, a guarantee of EUR 35 million intended to cover the credit risk associated with certain exposures of Tercas and, third, a guarantee of EUR 30 million intended to cover the costs arising from the tax treatment of the first measure.

4 During 2012, further to irregularities identified by the Bank of Italy, the Italian Government placed Tercas under special administration. Following the intervention by the FITD, Tercas was taken out of special administration and a number of capital increases were effected, inter alia, by the appellant in order to re-establish Tercas’ economic situation.

5 T‑98/16, T‑196/16 and T‑198/16, EU:T:2019:167; ‘the Tercas judgment of the General Court’.

6 C‑425/19 P, EU:C:2021:154; ‘the Tercas judgment of the Court of Justice’.

7 First, with regard to the condition relating to the unlawfulness of the conduct alleged against the Commission, the General Court, whilst finding that Article 107(1) TFEU, upon which the appellant relied in the present case, had to be characterised as a ‘rule intended to confer rights on individuals’ (paragraphs 76 to 96 of the judgment under appeal), concluded that there had not been a sufficiently serious breach of that rule (paragraphs 101 to 125 of that judgment). Second, with regard to the condition relating to the existence of a causal link, the General Court concluded that the appellant had failed to prove the existence of a direct causal link between the alleged unlawful conduct of the Commission and the alleged damage (paragraphs 126 to 161 of the judgment).

8 The General Court did, however, hold that the action was time-barred in so far as concerned one of the heads of damage relied on by the appellant, that is to say, that related to the damage associated with the reduction in staff numbers (paragraph 61 of the judgment under appeal).

9 This was in fact, more specifically, an agreement concluded between the appellant and the trade unions concerning an incentive scheme to encourage employees to leave, the aim of which was to reduce the staff by 85 employees.

10 The five-year limitation period would therefore have expired when the appellant made its prior application to the Commission (see, in that regard, paragraphs 61 and 68 of the judgment under appeal).

11 According to the appellant, the damage in question was not immediately perceptible on conclusion of the incentive scheme (on 30 December 2015), since, on that date, it was only possible to estimate the amount of the costs to be borne (for which funds had been earmarked in advance in the course of the 2015 financial year), whereas the disbursements had been made at a later date, that is to say, as employees signed up to that scheme offered by the appellant. Contrary to the view taken by the Commission, that ground is, in my view, admissible, since the appellant is not seeking a new assessment of the facts and the evidence by the General Court (without alleging a distortion of the facts), but is rather challenging the legal characterisation of those facts, as the General Court essentially characterised the alleged damage as ‘instantaneous’ but not ‘continuous’.

12 As the General Court observes in paragraph 29 of the judgment under appeal, that limitation period has the function, first, of ensuring protection of the rights of the aggrieved person, who must have sufficient time in which to gather the appropriate information with a view to a possible action, and, second, of preventing that person from being able to delay indefinitely the exercise of his or her right to damages. The General Court cited, to that effect, the judgments of 8 November 2012, Evropaïki Dynamiki v Commission (C‑469/11 P, EU:C:2012:705, paragraph 33 and the case-law cited), and of 7 July 2021, Bateni v Council (T‑455/17, EU:T:2021:411, paragraph 62 and the case-law cited).

13 See, inter alia, judgments of 19 April 2007, Holcim (Deutschland) v Commission (C‑282/05 P, EU:C:2007:226, paragraph 29), and of 8 November 2012, Evropaïki Dynamiki v Commission (C‑469/11 P, EU:C:2012:705, paragraph 34 and the case-law cited).

14 See, to that effect, judgments of 19 April 2007, Holcim (Deutschland) v Commission (C‑282/05 P, EU:C:2007:226, paragraph 30), and of 28 February 2013, Inalca and Cremonini v Commission (C‑460/09 P, EU:C:2013:111, paragraph 60).

15 In its action at first instance, the appellant acknowledged, in essence, that the costs relating to that scheme were ‘valued at a particular time and in advance based on the 2015 financial year’, whereas they were ‘actually borne in the course of subsequent years’.

16 Furthermore, the appellant did not experience any difficulties when quantifying that damage on the basis of an amount earmarked as at 31 December 2015.

17 See, to that effect, judgment of 28 February 2013, Inalca and Cremonini v Commission (C‑460/09 P, EU:C:2013:111, paragraph 80 and the case-law cited). See also Baranowska, N., and Machnikowski, P., ‘European Union’, Prescription in tort law: analytical and comparative perspectives, Intersentia, Cambridge, 2020, p. 706, and Dimitrova, Y., and Picod, F., ‘La prescription dans l’action en responsabilité contre l’Union européenne’, Revue des affaires européennes, 2021, No 4, p. 858.

18 See, inter alia, Opinion of Advocate General Jääskinen in Inalca and Cremonini v Commission (C‑460/09 P, EU:C:2012:632, point 45 and the case-law cited).

19 Furthermore, I note a contradiction in the arguments put forward by the appellant, which, after claiming that the damage at issue could not be calculated but only estimated when the incentive scheme was concluded, asserts that the alleged damage began to produce its effects only from the first disbursement, that is to say, from a date on which that damage likewise could not be calculated definitively.

20 According to the General Court, the unlawfulness of the Tercas decision arose from a conceptual error resulting from confusion between the condition relating to the imputability of aid and that relating to State resources, as well as from the Commission’s failure to set out and substantiate sufficient evidence capable of establishing that the measure under consideration could be imputed to the State (see, in that regard, paragraph 115 of the judgment under appeal).

21 See paragraphs 114 to 123 of the judgment under appeal.

22 See judgment of 11 January 2024, Dyson and Others v Commission (C‑122/22 P, EU:C:2024:11, paragraph 48 and the case-law cited; ‘the judgment in Dyson’). The requirement of a sufficiently serious breach of EU law seeks, whatever the unlawful nature of the measure in question, to avoid the situation where the risk of having to bear the losses alleged by the individuals concerned hinders the ability of the institution concerned to exercise fully its competence in the general interest, both in the context of its activities that are regulatory or involve economic policy choices and in the sphere of its administrative competence, without thereby leaving individuals to bear the consequences of flagrant and inexcusable misconduct (see, to that effect, judgment of the General Court of 3 March 2010, Artegodan v Commission, T‑429/05, EU:T:2010:60, paragraph 55 and the case-law cited).

23 See the judgment in Dyson (paragraph 49 and the case-law cited).

24 See the judgment in Dyson (paragraphs 50 and 51 and the case-law cited).

25 See the judgment in Dyson (paragraphs 52 and 53 and the case-law cited). See also Wildemeersch, J., ‘Section 2 – Les conditions du recours en responsabilité’, Contentieux de la légalité des actes de l’Union européenne, Bruylant, Brussels, 2019, pp. 466 to 479, and Nicolaides P., ‘The discretion of competition authorities and their liability for damage they cause with erroneous decisions’, European Competition Law Review, Vol. 45, No 7, Sweet & Maxwell, 2024, pp. 308 to 316.

26 See, to that effect, Lenaerts, K., et al., EU Procedural Law, 2^nd ed., Oxford University Press, 2023, p. 501. According to those authors, the liability test applied by the European Union judicature contains a subjective element linked to the discretion enjoyed by the institution, body, office or agency, together with the consideration of factors such as, inter alia, the complexity of the situations to be regulated, the difficulties in the application or interpretation of the legislation, the degree of clarity and precision of the rule of EU law breached, and whether the error was inexcusable or intentional.

27 For example, the Court has established that the mere infringement of EU law could suffice to establish the existence of a sufficiently serious breach of that law capable of giving rise to the European Union’s non-contractual liability in a situation in which the Commission – whilst being required to repay to an undertaking the amount of a fine provisionally paid, together with default interest, following the annulment of a decision by which it had found that that undertaking had infringed Article 101 TFEU – had failed to pay such interest (see, to that effect, judgment of 20 January 2021, Commission v Printeos, C‑301/19 P, EU:C:2021:39, paragraphs 104 and 105).

28 See the judgment in Dyson (paragraph 54). In other words, there is no automatic link between the lack of discretion of the institution concerned and the classification of a ‘sufficiently serious breach of EU law’. Although the extent of the discretion of the institution concerned is a determining factor, it is not an exclusive criterion (see, to that effect, judgment of the General Court of 3 March 2010, Artegodan v Commission, T‑429/05, EU:T:2010:60, paragraphs 59 and 60 and the case-law cited).

29 In paragraph 70 of the Tercas judgment of the General Court, the General Court essentially criticised the Commission for having not sought to make a clear distinction, in the case in question, between the imputability of the measure to the State and the question whether that measure had been granted through State resources, which are separate and cumulative conditions. In its Tercas judgment, the Court of Justice did not rule on that aspect, but rather simply examined the condition of the imputability of the measures at issue to the Italian authorities (see paragraph 64 of that judgment).

30 It is apparent from paragraphs 87 to 90 of the Tercas judgment of the General Court, as confirmed by paragraph 67 of the Tercas judgment of the Court of Justice, that the Commission could not rely on the unlikelihood of an absence of actual influence and control by the public authorities over the private entity providing the aid, but was required to set out and substantiate sufficient evidence capable of establishing that the aid measure under consideration had been adopted under the actual influence or control of the public authorities. In essence, the Commission did not take into account the absence of a ‘link of a capital nature’ between the entity concerned and the State (see paragraph 73 of the Tercas judgment of the Court of Justice).

31 As follows from the case-law of the Court, the list of the constituent elements of State aid within the meaning of Article 107(1) TFEU itself varies depending on whether, according to the circumstances, some of those elements are assessed jointly. On the basis of the most detailed list, the following elements may be distinguished: the imputability of the measure to the State and the use of State resources (see, inter alia, judgment of 2 March 2021, Commission v Italy and Others, C‑425/19 P, EU:C:2021:154, paragraph 58 and the case-law cited); the economic advantage for the recipient (see, inter alia, judgment of 24 July 2003, Altmark Trans and Regierungspräsidium Magdeburg, C‑280/00, EU:C:2003:415, paragraph 84 and the case-law cited); the selectivity of the measure (see, inter alia, judgment of 10 September 2024, Commission v Ireland and Others, C‑465/20 P, EU:C:2024:724, paragraph 75 and the case-law cited); the distortion of competition (see, inter alia, judgment of 19 September 2000, Germany v Commission, C‑156/98, EU:C:2000:467, paragraph 33 and the case-law cited); and the effect on trade between Member States (see, inter alia, judgment of 7 March 2024, Fallimento Esperia and GSE, C‑558/22, EU:C:2024:209, paragraph 64 and the case-law cited).

32 See, inter alia, judgment of 21 December 2016, Commission v World Duty Free Group and Others (C‑20/15 P and C‑21/15 P, EU:C:2016:981, paragraph 53 and the case-law cited).

33 See, inter alia, judgment of 27 June 2017, Congregación de Escuelas Pías Provincia Betania (C‑74/16, EU:C:2017:496, paragraph 78).

34 For example, in the Tercas judgment of the Court of Justice, the Court clarified, in its summary of case-law, that classification as ‘State aid’ for the purposes of Article 107(1) TFEU requires four conditions to be satisfied, the first of which is that ‘there is intervention by the State or through State resources’ (see paragraph 57 of that judgment).

35 See, respectively, paragraphs 87 to 89 of the Tercas judgment of the General Court and paragraphs 65 to 67 of the Tercas judgment of the Court of Justice.

36 According to the Commission’s definition, the FITD was a private entity which had a very close relationship with the public authorities, in particular on account of its formation, its mission, its organisation, its origin and the development of the measure (see, to that effect, paragraphs 121 to 131 of the Tercas decision).

37 That is the case, in particular, where the ‘private operator principle’ is applied in the context of a review of the condition of the advantage for the recipient (see, for example, judgment of 11 November 2021, Autostrada Wielkopolska v Commission and Poland, C‑933/19 P, EU:C:2021:905, paragraph 116 and the case-law cited).

38 See, by way of illustration, Merola, M., and Caliento, F., ‘Is the notion of aid broadening or shrinking over time, and if so, why? A subjective view on the rationale of the case law’, EU State Aid Law, Edward Elgar Publishing, Cheltenham, 2020, pp. 18 to 53. Those authors identify several examples that underscore the difficulties inherent in the definition of State aid, inter alia, as far as concerns the concept of ‘State resources’ (see, in particular, pp. 26 to 31) and also point to a recent upsurge of cases in which the Court of Justice has set aside the judgments of the General Court.

39 It is should be noted, in that regard, that the Tercas judgment of the Court of Justice was given by the Court sitting in a Grand Chamber.

40 See points 23 and 24 of this Opinion. In such situations, the requirement of a sufficiently serious breach of EU law would be rendered redundant if any infringement committed by the Commission, like any public administration acting within its areas of expertise, was sufficient to satisfy that requirement.

41 To reproduce the observations made by the Commission at the hearing, it is a situation in which it committed an error (and, in that regard, issued a ‘mea culpa’), but one which, however, does not go beyond what is expected of a diligent administrative authority (in which, in such circumstances, it should have issued a ‘mea maxima culpa’).

42 See paragraphs 132 to 161 of the judgment under appeal.

43 See, inter alia, judgment of 27 April 2023, Fondazione Cassa di Risparmio di Pesaro and Others v Commission (C‑549/21 P, EU:C:2023:340, paragraph 114 and the case-law cited). See also, to that effect, Lenaerts, K., et al., EU Procedural Law, 2^nd ed., Oxford University Press, 2023, pp. 512 to 514. As those authors observe, there is no standard form of words in the case-law of the Court as regards that requirement, which is expressed in different ways by the Courts of the European Union, with the primary underlying idea being that the European Union is liable only for damages which are the direct consequence of the unlawful measures or conduct of its institutions, bodies, offices and agencies (see also Wildemeersch, J., ‘Section 2 – Les conditions du recours en responsabilité’, Contentieux de la légalité des actes de l’Union européenne, Bruylant, Brussels, 2019, pp. 484 to 487). In addition, some authors draw a distinction between, on the one hand, the concept of ‘factual causality’, which answers the question whether the unlawful conduct is a ‘conditio sine qua non’ for the damage claimed, and, on the other hand, ‘legal causality’, which imposes additional requirements so as to exclude the liability of EU institutions or bodies for damage that is too remote by requiring that the damage flows sufficiently directly from the unlawful conduct, that there is a sufficiently direct nexus between the conduct and the damage, or that the conduct is the determining cause of the damage (see, to that effect, Biondi, A., and Farley, M., The right to damages in European law, Kluwer Law International, 2009, pp. 146 to 154). That same distinction was reproduced in the Opinion of Advocate General Ćapeta in WS and Others v Frontex (Joint return operation) (C‑679/23 P, EU:C:2025:427, points 101 to 109). See also, for comparative studies of the different approaches to causality in various EU States, Infantino, M., and Zervogianni, E., ‘The European Ways to Causation’, Causation in European Tort Law, Cambridge University Press, 2017, pp. 85 to 128, and Bussani, M., Sebok, A., J., and Infantino, M., ‘Causation’, Common Law and Civil Law Perspectives on Tort Law, Oxford Academic, online edition, 2022, pp. 177 to 212.

44 See paragraph 137 of the judgment under appeal.

45 See paragraphs 138 to 141 of the judgment under appeal.

46 Judgment of 27 April 2023 (C‑549/21 P, EU:C:2023:340). The appellant observes, inter alia, that, unlike the case which formed the subject of that judgment, in which the Commission exerted pressure on the Italian authorities which did not relate to a specific measure, in the present case the Commission adopted a binding decision.

47 See, to that effect, judgment of 18 March 2010, Trubowest Handel and Makarov v Council and Commission (C‑419/08 P, EU:C:2010:147, paragraph 59 and the case-law cited). The concept of a ‘sufficiently direct link’ is interpreted narrowly by the case-law, as is confirmed, inter alia, by the General Court, where it held that the mere fact that the allegedly unlawful conduct of the EU institution or body concerned constituted a necessary condition (a sine qua non) for the damage, in the sense that it would not have occurred without that conduct, is not sufficient to establish a sufficiently direct causal link under that case-law (see, inter alia, to that effect, judgment of 23 May 2019, Remag Metallhandel and Jaschinsky v Commission, T‑631/16, EU:T:2019:352, paragraph 52 and the case-law cited). See also Havu, K., and Kurki-Suonio, S., ‘Damages Liability of the EU for Harm Caused by Excessive Duration of Court Proceedings’, European Public Law, Vol. 27, No 2, Kluwer Law International BV, 2021, pp. 305 to 330.

48 See, inter alia, to that effect, judgment of 16 July 2009, Commission v Schneider Electric (C‑440/07 P, EU:C:2009:459, paragraph 192).

49 See, inter alia, to that effect, judgment of 19 October 2017, Agriconsulting Europe v Commission (C‑198/16 P, EU:C:2017:784, paragraph 22).

50 See, inter alia, to that effect, judgment of 1 August 2025, France and Commission v CWS Powder Coatings and Others (C‑71/23 P and C‑82/23 P, EU:C:2025:601, paragraph 67 and the case-law cited).

51 By contrast, the appellant does not contest the General Court’s assessment, laid out in paragraph 148 of the judgment under appeal, that the appellant failed to explain why the Tercas decision prevented it from attracting new customers, even though the voluntary intervention (which replaced the intervention of the FITD for the benefit of Tercas and was not authorised by that decision) had already been decided upon two months after that decision. However, I doubt that that element alone can form the basis of the General Court’s assessment regarding the concomitant causes of the alleged damage.

52 In so doing, the appellant itself acknowledges the existence of an additional factor liable to have contributed to the alleged damage.

53 See paragraph 144 of the judgment under appeal.

54 See paragraph 146 of the judgment under appeal.

55 See paragraph 153 of the judgment under appeal.

56 See paragraph 145 of the judgment under appeal.

57 See paragraph 146 of the judgment under appeal.

58 See paragraph 149 of the judgment under appeal.

59 See paragraphs 156 and 158 of the judgment under appeal.

Full text

Continue your research in ChatGPT or Claude