On 10 December 2025, the EU General Court (the Court) ruled on Intel’s latest appeal against fines imposed on it by the European Commission for abusing its dominant position on the global market for computer CPUs. As explored in this article, this latest ruling may finally bring to a close proceedings that have now lasted over 25 years.

Intel’s appeal related to a September 2023 Commission decision, in which it reimposed a fine of €376m on Intel for engaging in so-called “naked restrictions”, in the form of payments to PC manufacturers in return for them cancelling or postponing the sale of computers incorporating CPUs from rival chip manufacturer AMD.

The Court rejected most of the arguments raised by Intel, concluding that the Commission had jurisdiction to impose the fine and that Intel’s procedural rights had not been infringed. However, the Court agreed that the penalty did not properly reflect the reduced duration and gravity of the infringement committed by Intel. It therefore reduced the amount of the penalty to €237m.

Background

Following a complaint lodged with the Commission by AMD in 2000, the Commission investigated Intel’s conduct regarding the worldwide supply of CPUs incorporating the “x86” chip architecture. In 2009 it imposed a fine of €1.06bn on Intel for engaging in two types of exclusionary conduct: (i) loyalty-inducing rebates – incentivising four PC manufacturers/retailers to exclusively sell computers featuring Intel CPUs; and (ii) naked restrictions vis-à-vis three PC manufacturers, in breach of Article 102 of the Treaty on the Functioning of the European Union (TFEU).

Intel appealed. The decision was initially upheld by the General Court, but in 2017 the EU Court of Justice (CJEU) ruled that the Court had erred in its treatment of Intel’s rebates and referred the case back to it for reconsideration. The General Court, applying criteria set out in the CJEU’s judgment, annulled the Commission decision in 2022 insofar as it related to the loyalty rebates. A subsequent appeal by the Commission was then rejected by the CJEU in October 2024, and the €1.06bn fine definitively quashed. See our previous article on the CJEU decision for more details.

The 2022 General Court ruling did not, however, overturn the finding of infringement in respect of the naked restrictions. The Commission therefore issued a further (very short) decision, in which it imposed a new, lower fine. Crucially, there was no new finding of infringement in that decision – the Commission simply calculated a new fine in respect of the infringement it found in 2009, save that the scope, duration and gravity of the infringement were reduced as a result of the Court’s 2022 ruling. Intel appealed that further decision.

The Court’s ruling

Intel requested that the September 2023 decision be quashed, or that the fine imposed therein be annulled or reduced, broadly relying on three heads of argument:

1. Lack of jurisdiction

Intel submitted that the Commission lacked the jurisdiction to impose a new fining decision in respect of the naked restrictions alone. This was on the basis that, taken on its own, that conduct was radically different to the “single continuous infringement”¹ found by the Commission in 2009 (which also incorporated the loyalty rebates) and its effects in the EEA had not sufficiently been established (two of the three naked restrictions were implemented outside the EEA).

The Court, however, noted that the finding of infringement in respect of the naked restrictions had become definitive, as Intel had not appealed the 2022 ruling. Moreover, Intel had previously withdrawn its jurisdictional arguments against the 2009 fining decision, and neither the Court nor the CJEU had held that the naked restrictions could no longer form part of the single and continuous infringement originally found by the Commission. There was, therefore, no need for the Commission to re-establish its jurisdiction to impose the revised fine – in doing so it was simply complying with the Court’s 2022 judgment.

2. Failure to respect Intel’s procedural rights

Intel argued that the September 2023 decision did not include adequate reasoning; that rather than sending Intel a letter and then issuing a new fining decision, the Commission should instead have issued a new Statement of Objections; and that Intel’s right to be heard had not, therefore, been observed by the Commission.

The Court disagreed. It considered that Intel’s arguments were based on the false premise that the Commission had defined a new, single and continuous infringement in its September 2023 decision. Moreover, Intel had a clear understanding of the Commission’s reasoning and the factual basis for the new fine. Given no new infringement had been found by the Commission in its September 2023 decision, there was no need for it to issue a new Statement of Objections. And by first sending a letter informing Intel of its intention to impose a new fine, as well as the basis and calculation method for that fine, and inviting Intel to submit its observations, the Commission had respected Intel’s right to be heard.

3. The level of the new fine was disproportionate and unlawful

Intel put forward multiple arguments in this regard. Principally these concerned the significance of the naked restrictions as compared to the overall infringement originally found by the Commission. For example, Intel noted that the new fine was almost three-eighths of the original penalty, despite the naked restrictions affecting only 8% of its sales over the relevant period and being in place for only 19% of the duration of the single and continuous infringement found in the 2009 decision. It also pointed to the fact that the “gravity” multiplier applied when calculating the new fine (which reflects the seriousness of the infringement) was only 20% lower than that originally applied.

The Court declined to strike out the new fine on this basis. The Commission had properly taken into account the reduced duration of the infringement. And its approach to calculating the new fine – including the value of sales taken into account – was in line with its 2006 guidelines on calculating antitrust penalties. The prolonged nature of the proceedings also did not amount to a mitigating factor that the Commission should have taken into account. Therefore, the new fine was not disproportionate and unlawful.

Nevertheless, the Court – exercising its unlimited jurisdiction – decided to reduce the amount of the penalty by 37%, to reflect more equitably the gravity and duration of the remaining infringement. In particular, the Court pointed to the fact that there was a more than 12-month gap between two of the naked restrictions, and the restrictions affected only a relatively modest number of computers – neither of which had, in its view, been fully factored in by the Commission.

The end of the saga?

This latest ruling could well bring to a close one of the most convoluted and long-running sagas in the history of EU antitrust enforcement. Whilst the Commission might still choose to appeal the €139m reduction in the fine, it may consider it best to move on and draw a line under the proceedings. The real damage to its enforcement efforts was done in the CJEU’s 2017 and 2024 rulings, as a result of which the Commission is, effectively, required to conduct an econometric “as-efficient competitor” test in order to establish that a given loyalty rebate scheme is exclusionary and therefore abusive (unless the firm under investigation puts up no defence to the allegations against it).

In reality, the stakes for the Commission in this latest appeal were not that high, as the unlawfulness of the naked restrictions had already become definitive (a point noted by the Commission in its draft Article 102 guidelines, in which the concept of a naked restriction – conduct that has no explanation other than to exclude competitors from the market, and can therefore be presumed to be exclusionary – featured prominently). And had Intel’s arguments gained more traction with the Court, it would have been open to the Commission to reimpose a new fine.

As to Intel, it argued before the Court that its reputation has been damaged by more than two decades of proceedings concerning pricing practices that were wrongly regarded as anticompetitive. Nevertheless, it can console itself with the fact that, at the end of those proceedings, it has secured a more than 75% reduction in a penalty that was, at the time, the largest ever imposed by the Commission for a breach of EU competition law.

Key takeaways

The Intel case, along with the Air Cargo cases, appears to have prompted a degree of soul-searching within the EU judiciary. It is understood that the CJEU is keen to reduce the likelihood of similar sagas, marked by lengthy procedural “ping-pong” between itself, the General Court and the Commission, in the future. To that end, the CJEU, as the highest court, has signalled that it is inclined to resolve disputes on the merits at the appeal stage more often, rather than remitting cases back to the General Court.

For businesses and legal practitioners, this shift in judicial practice carries strategic implications:

• Front-load the strategy: litigants must ensure their initial appeals to the General Court are exhaustive and robust, as the opportunity to "re-argue the case" via referral back from the CJEU is becoming less likely.
   
• Precision in the CJEU: when drafting an appeal to the CJEU, parties must be acutely aware that it may choose to deliver a definitive ruling rather than simply identifying a point of law and sending it back.
   
• Strategic cross-appeals: parties should carefully consider the appropriateness of cross-appeals to protect their position, ensuring all potential outcomes are addressed while the case is before the highest court.

While Intel ultimately secured a very significant reduction in its original penalty and changed the requirements for enforcing abuse of dominance cases, the judicial legacy of this case might also include a more streamlined, "one-and-done" approach to EU competition litigation.