A Four-Year Fight Over One Word: "Knowledge"
On 16 July 2026, the Court of Justice of the European Union ruled in Case C-421/24 (Google Ireland v AGCOM) that Google cannot automatically claim the hosting-liability exemption under Article 14 of the e-Commerce Directive for YouTube videos uploaded by a creator with whom it has a commercial partnership. The case traces back to 19 July 2022, when Italy's communications regulator AGCOM fined Google Ireland €750,000 and ordered the removal of roughly 630 videos from a creator's channels that promoted gambling sites and invited viewers to share their own winnings — conduct AGCOM said violated Article 9 of the 2018 "Decreto Dignità," Italy's near-total ban on gambling advertising (AGCOM's legal basis).
The Court's reasoning turns on a narrow but consequential fact: before signing the creator into YouTube's Partner Programme, Google had reviewed the channel's themes, its most-viewed and newest videos, and associated metadata (Curia press release No. 109/26). The Court held that this vetting — done to assess brand safety and revenue-sharing eligibility — gave Google "specific knowledge" of the content sufficient to defeat the passive-intermediary status Article 14 requires. Hosting gambling advertising, the Court clarified, still falls within the Directive's scope even though gambling activity itself is excluded from EU services law (Il Sole 24 Ore). The case now returns to Italy's Council of State, which must determine whether Google "could not reasonably have been unaware" of the channels' gambling focus given what it had already reviewed — effectively deciding whether the original €750,000 fine stands (gamingtechlaw.com).
The Case for the Ruling
AGCOM's position deserves a fair hearing before it gets a rebuttal. Italy's gambling-ad ban exists because gambling disorder is a real public-health harm, and the Decreto Dignità was written broadly and passed by Parliament precisely to close loopholes that direct-advertising bans leave open — sponsorships, influencer content, algorithmically boosted clips of other people's winnings. If Google's YouTube Partner Programme actively reviews a channel's content before agreeing to split ad revenue with it, that review is not incidental; it is due diligence performed for Google's own commercial benefit. A platform that profits from a partnership it vetted, the argument goes, should not get to disclaim knowledge of what it vetted. Regulators have legitimate reason to worry that revenue-sharing arrangements are where "neutral hosting" quietly becomes curated publishing, and that platforms could otherwise structure creator programs specifically to launder editorial control through willful blindness.
Where the Reasoning Overreaches
The problem is that the Court has now made the scope of a platform's liability shield turn on the intensity of its safety and quality review — which inverts the incentive the DSA and the e-Commerce Directive were built to protect. Both instruments deliberately shield platforms that act in good faith to review content, precisely so they are not punished for trying. If examining a channel's themes and metadata before a monetization decision counts as acquiring disqualifying "knowledge," the rational response for any platform is to review less, not more — the opposite of what regulators, including AGCOM, actually want from content moderation. As EFF has warned in its analysis of this same line of CJEU cases, treating ordinary platform-side organization and vetting as equivalent to editorial control risks "deterring proactive moderation" across the board, not just in gambling (EFF).
This also creates a monetization-specific liability trap that has nothing to do with the DSA's actual due-diligence exemption, which explicitly allows voluntary investigations without forfeiting hosting status. A revenue-sharing decision is a business judgment about ad-quality and brand safety, not a claim that the platform read and approved every future upload. Collapsing those two things means any platform running a creator fund, ad-share program, or influencer marketplace — in any sector, not just gambling — now faces a live legal question every time it does routine partner vetting: did that review create constructive knowledge of everything the partner posts afterward? For a company the size of Google that can absorb a €750,000 fine and years of litigation, this is an annoyance. For the mid-sized European platforms the DSA claims to want to nurture, the safer and cheaper answer is to stop vetting partners at all, or to stop running revenue-share programs in jurisdictions with aggressive sectoral ad bans.
What Should Happen Next
The Council of State's remand is the right venue to test this narrowly: did Google, in this specific case, actually know these particular channels were gambling-promotion vehicles, or did it merely conduct generic partner-eligibility review? If the former, AGCOM's fine is a clean win against knowing facilitation, no CJEU doctrine required. If the latter, upholding the fine would functionally require every platform ad-partnership program in the EU to choose between abandoning content review or accepting open-ended sectoral liability — a cost the DSA's own hosting-exemption architecture was designed to avoid. Proportionality here means judging what Google actually knew about these specific channels, not treating routine partner vetting as a universal admission of publisher liability.