UK Trademark Cases 2025-26: Umbro Lost, Thatchers Won
Truviss monitors UK marketplaces and social platforms continuously, so counterfeit activity gets flagged while it is happening, not years later in a judgment.
Book a demo- Five UK trademark and counterfeit rulings landed between January 2025 and May 2026, and brand owners won barely half of them.
- Umbro’s parent Iconix lost outright at the Supreme Court. Thatchers won on appeal against Aldi. Getty Images won only a sliver against Stability AI. ShortsTV lost outright against Google. An influencer selling counterfeit luxury goods lost outright at IPEC.
- None of these turned on brand size or fame. Each turned on a narrow legal question: actual confusion, unfair advantage, distinctiveness, or straightforward counterfeit sale.
- The one genuine counterfeit case took months to reach judgment. The trademark-scope disputes took years and multiple appeals. Litigation is a backstop, not a detection system.
The scoreline: brand owners won two, lost two, and split one
Five UK court rulings between January 2025 and May 2026 put trademark and counterfeit law properly to the test, and the results do not follow a pattern anyone selling “brand protection” would want to pretend they do. Umbro’s owner lost at the UK Supreme Court. Thatchers beat Aldi at the Court of Appeal. Getty Images won against Stability AI, but the judge called the win “extremely limited in scope.” ShortsTV lost against Google, also at the Court of Appeal. And a UK influencer selling counterfeit Fendi, Loewe, Dior and Celine goods lost outright at the Intellectual Property Enterprise Court.
Two clear brand-owner wins. Two clear losses. One partial win that reads more like a loss in practice. This is not a story about which side had better lawyers. It is a story about what UK courts are actually looking for when a brand claims someone else has crossed the line, and it is worth reading properly before assuming a famous logo or a big legal budget settles anything.
Iconix (Umbro) v Dream Pairs [2025] UKSC 25: the brand owner lost
On 24 June 2025, the UK Supreme Court unanimously allowed an appeal by Dream Pairs Europe, restoring the original trial judge’s finding that Dream Pairs’ football boots did not infringe Umbro’s double-diamond logo. Iconix Luxembourg Holdings, which owns the Umbro trademark, had won at the Court of Appeal after losing at first instance. The Supreme Court reversed that again, back in Dream Pairs’ favour.
The legal question was narrow and specific: post-sale confusion under section 10(2) of the Trade Marks Act 1994, whether someone seeing the boots after purchase, not at the point of sale, might mistake them for Umbro’s. The trial judge had already found no likelihood of confusion. The Court of Appeal disagreed and substituted its own assessment. The Supreme Court’s point was procedural as much as substantive: an appellate court does not get to swap in its own multi-factor judgment call over a trial judge’s factual finding just because it would have weighed the evidence differently. The original no-infringement finding stood.
This case belongs at the top of any UK trademark roundup precisely because it is not a brand-owner win. A household-name logo, a recognisable design similarity, and a well-resourced claimant still lost, because the trial evidence did not support a confusion finding and the appeal courts are not supposed to relitigate facts.
Thatchers v Aldi [2025] EWCA Civ 5: the brand owner won on appeal
Three weeks earlier, on 20 January 2025, the Court of Appeal reached the opposite kind of result for a different claimant. Thatchers Cider had lost at the Intellectual Property Enterprise Court, which found no infringement in Aldi’s Taurus Cloudy Lemon Cider packaging. The Court of Appeal overturned that decision and found Aldi had taken unfair advantage of Thatchers’ trade mark under section 10(3), the broader protection that does not require proof of consumer confusion, only that an unfair link was drawn in a shopper’s mind between the two products.
Put this next to Umbro and the contrast is direct: two established brands, both disputes reaching the Court of Appeal within months of each other, one flipped toward the brand owner and one flipped away. The difference was not the courts changing their approach to trademark law. It was the underlying evidence: Aldi’s packaging design similarity and the specific unfair-advantage case Thatchers built were strong enough to succeed under section 10(3) even where a straightforward confusion argument might not have. Aldi has reportedly signalled it will seek permission to appeal further to the Supreme Court, so treat this outcome as significant but not necessarily final.
Getty Images v Stability AI [2025] EWHC 2863 (Ch): a win narrow enough to feel like a loss
Getty Images sued Stability AI in the UK High Court over AI image generation, and the judgment landed on 4 November 2025 from Mrs Justice Joanna Smith. It is worth being precise about what this case actually is before using it in any brand-protection context: this is primarily an AI training-data and copyright dispute, not a counterfeit case. Getty’s own primary copyright claims were abandoned partway through trial, and a secondary copyright claim was rejected.
On trademark specifically, the court dismissed Getty’s broader section 10(3) claim entirely and found only “extremely limited” infringement under sections 10(1) and 10(2), confined to early versions of Stable Diffusion where Getty’s own watermark occasionally appeared, distorted, in AI-generated outputs. That is the entire scope of the trademark win. It belongs in this roundup as the clearest illustration of a brand technically prevailing on a narrow point while losing almost everything else it actually wanted, which functions as a loss in every practical sense.
Shorts International v Google [2026] EWCA Civ 668: the brand owner lost again
On 25 May 2026, the Court of Appeal dismissed an appeal by Shorts International (SIL), confirming that Google’s “YouTube Shorts” branding did not infringe SIL’s trademarks. The court went further and found one of SIL’s own registrations, the word mark SHORTSTV, invalid for lacking distinctiveness in the first place.
This ties directly back to the throughline: descriptive or weakly distinctive marks are difficult to win on even against a household-name defendant with obvious commercial scale. Courts keep testing whether the claimed mark was ever strong enough to protect on its own terms, before they ever get to whether the accused use looks similar to it.
Fendi Italia SRL v Rolo Fashion Ltd (IPEC, 2026): the actual counterfeit case in the set
The four cases above are all disputes over the scope of legitimate trademark protection between established, legally operating companies. This one is different in kind. Georgia Aldridge, a UK-based influencer with roughly 32,000 Instagram followers who ran Sloane House Marketing and a side-hustle dropshipping operation, Rolo Fashion, was found by the Intellectual Property Enterprise Court to have sold counterfeit Fendi, Loewe, Christian Dior and Celine goods sourced from AliExpress, so-called “superfakes,” through her online store. IPEC ordered her to pay £213,000, covering lost profits from an estimated 713 sales plus licensing income, and rejected a separate reputational-damage claim from the brands.
Unlike the trademark-scope disputes above, there was no argument here about how similar a logo needed to be or whether a mark was distinctive enough. This was straightforward sale of counterfeit goods, caught and financially quantified after the fact through litigation. By the time judgment landed, hundreds of individual sales had already gone through.
See how Truviss surfaces counterfeit sellers across marketplaces and social platforms before a case ever needs to go to court.
Explore Marketplace Scanner UKWhat actually decided these outcomes
Line the five cases up and none of them turned on brand fame, market size, or which side looked like it “should” obviously win.
- Umbro lost because the trial judge’s factual finding on post-sale confusion was not open to appellate revision.
- Thatchers won because the specific evidence of unfair advantage under section 10(3) held up on appeal even after an earlier confusion-based loss.
- Getty won almost nothing because its actual proven infringement was confined to a narrow historical slice of the defendant’s product.
- ShortsTV lost because its own mark’s distinctiveness did not survive scrutiny.
- Sloane House lost outright because the underlying conduct was straightforward counterfeit sale rather than a scope dispute, and the evidence of actual sales volume was concrete.
The practical read for a brand team: outcomes here increasingly hinge on how specific and well-evidenced the case is, actual confusion, actual unfair advantage, actual distinctiveness, actual sales volume, rather than on how well known the brand is walking in. This is the same pattern behind online brand abuse: the strength of a claim tends to come down to documented, real-world evidence rather than the underlying brand’s reputation.
It also echoes a related pattern on the other side of the Atlantic: a Supreme Court ruling narrowing what counterfeiters actually have to pay showed that even a favourable judgment does not always translate into the recovery a brand expects. Winning is not the same as collecting, and here, winning is not even guaranteed.
What this means for brand protection strategy
Timelines matter as much as outcomes. Sloane House reached judgment in months. Umbro and Thatchers each took years and multiple appeal stages to resolve, and even Thatchers’ win may not be final. Getty’s “win” arrived nearly two years after Stability AI’s alleged use began, covering only a fraction of what Getty originally claimed.
The Sloane House case is the clearest argument for catching activity before it reaches a courtroom at all. By the time IPEC quantified 713 counterfeit sales, all of them had already happened and the brands involved had already lost that revenue. Continuous monitoring across marketplaces and social platforms, the kind Truviss’s Marketplace Scanner and social monitoring modules run for UK brands, is what surfaces seller activity like this while it is still active rather than after a judge has finished totting up the damage. This is not a substitute for litigation when litigation is warranted; it is the earlier layer that gives a brand the option to act before hundreds of sales become a fait accompli.
The same logic applies to the lookalike-packaging pattern behind Thatchers v Aldi, which sits in the same territory as the ongoing Mondelez v Aldi dupe-packaging dispute in the US: courts will eventually rule on any individual case, but a brand’s day-to-day exposure to copycat listings and packaging keeps moving in the meantime, on marketplaces and social platforms a single lawsuit was never built to watch continuously. Wider digital risk protection sits alongside legal enforcement rather than replacing it, catching the activity a court will only ever see months or years after the fact.







