On 12 November 2026, Advocate General Juliane Kokott told the Court of Justice that the General Court erred in law when it annulled the Commission's €1.49 billion AdSense fine against Google. She proposed that the 2024 judgment be set aside and the case sent back to the General Court (Case C-826/24 P, Commission v Google and Alphabet). The opinion is not binding. The judges have begun deliberating, and a judgment is expected next year.
What the opinion says
In a March 2019 decision, the Commission found that Google had committed three infringements of Article 102 TFEU between 2006 and 2016, which together formed a single and continuous infringement. The conduct involved AdSense for Search, which lets websites with built-in search boxes show Google search ads for a share of the revenue. Google's model contracts from March 2009 contained three clauses. The exclusivity clause barred rival search ads on the sites concerned. The placement clause required a minimum number of Google ads and kept competing ads away from the most prominent positions. The prior-authorisation clause required Google's approval before a publisher changed how search ads were displayed. Court of Justice press release 164/26
The General Court annulled the decision on 18 September 2024. It reasoned that the Commission had examined the relevant circumstances inadequately, in particular the duration of the contractual relationships, and had therefore not demonstrated an infringement.
Kokott's point is about the standard of review, not the facts. The Court of Justice has recognised several criteria for judging whether contractual commitments by a dominant firm are exclusionary, known as the Intel criteria, and duration is one of them. In her view the General Court wrongly read the Commission's duty to examine all relevant circumstances as requiring a detailed analysis of every circumstance under each individual criterion. She said that this imposed "excessive requirements" on the examination of duration. She also said the General Court did not weigh duration against the other criteria in their overall context.
Kokott noted that the General Court had itself found that the clauses were in principle capable of producing the exclusionary effects the Commission identified. She said it was enough for the Commission to assess cumulatively the contracts significant to how competition works on the market, with their durations. Contract-by-contract dissection was not required. If the Court of Justice agrees, the General Court will have to revisit Google's other pleas, including its objections to the size of the fine. Kokott added that the fine should in any case be reduced, because findings that exclusionary effects were not shown for 2016 were not appealed by the Commission and are final.
The case for the Commission
The strongest argument for this reading is practical. Abuse cases against dominant platforms turn on the market-wide effects of many contracts. If the Commission must model every agreement separately, enforcement against standardised terms becomes nearly impossible, and the harm is exactly what the rules are meant to catch. Rival search-ad providers could not reach publishers' sites on equal terms for years. The Commission is also not the author of its own standard: the Intel framework came from the Court of Justice, and Kokott's reading arguably restores it rather than expanding it.
Why an innovation-friendly reader should still care
The pro-innovation worry is not that Google should escape scrutiny. It is that the standard of proof in Article 102 cases should stay predictable. Fines of this size shape how every platform drafts its commercial terms. Some of Google's AdSense clauses were dropped or amended in September 2016, within months of the Commission opening proceedings in July 2016, which suggests that targeted remedies can work quickly.
Three points follow from the opinion.
- The cumulative approach needs discipline. It is defensible where contracts are standardised and the Commission can show how the clauses worked in practice. It is weaker where average values stand in for facts about individual publishers. The Court should say how much empirical substantiation is enough.
- Partial wins matter. The unchallenged 2016 findings show that the effects analysis can fail for part of a period and that this is not treated as a technicality. A reduced fine would be a reasonable consequence of that, and it shows proportionality working inside the framework.
- Remand is not reinstatement. The opinion sends the case back. Google keeps the chance to argue the rest of its case, including the amount of the fine, before the General Court.
What to watch
The fine is one of six EU antitrust penalties that have cost Google €10.38 billion over nearly two decades, according to TNW's report. That is large enough to make the legal standard a matter of commercial policy, not only of doctrine. Kokott's opinion suggests the Court of Justice may hold that the General Court raised the bar for the Commission too high. If it does, expect more Article 102 decisions built on cumulative contract analysis.
The better outcome is a judgment that backs the cumulative method and also states what evidence the Commission must produce on effects and duration, so the next investigation is judged on evidence and not on how much effort was spent on exhaustive analysis. Advocates General's opinions are followed often but not always, and the Court may well write its own reasoning. Companies and regulators should read the judgment, not the opinion, as the guide.