It is not FRAND, despite the deserved attention paid to Panasonic v OPPO (Mannheim LD, 22 November 2024, UPC_CFI_210/2023) and Huawei v Netgear (Munich LD, 18 December 2024, UPC_CFI_9/2023). It is not inventive step, despite the Court of Appeal’s Amgen framework. It is jurisdiction. The CJEU’s BSH v Electrolux ruling has fortified the willingness of UPC divisions to adjudicate infringement of European patents well beyond the UPC’s own borders.
A pattern that began as a Düsseldorf experiment in January 2025 has become a routine feature of the docket. Fifteen months after BSH, 31 UPC decisions have already cited it. Mannheim alone accounts for more than half. This is the single most operationally consequential development of year three.
And it has surfaced a workflow problem that European patent teams are only beginning to address: defendants in long-arm UPC actions now have to mount validity arguments under the substantive patent law of each third State whose designation is in suit (UK, Swiss, Turkish, Spanish).
The Grand Chamber handed down BSH v Electrolux on 25 February 2025. The headline holding is short. Under Article 4(1) of the Brussels Ia Regulation, the court of the defendant’s domicile has jurisdiction to rule on infringement of a European patent in respect of all designations, including non-EU States such as the United Kingdom or Turkey. A nullity defence under Article 24(4) does not strip that jurisdiction over infringement. The court may also rule inter partes on third-State validity, provided it does not order any amendment of a foreign register.
For practitioners who lived through the 2006 GAT v LuK contraction of cross-border patent litigation, this is a meaningful reopening. It removes the principal obstacle that had pushed cross-border patent disputes back into national silos for nearly two decades.
The judgment lands particularly heavily on the UPC. Every defendant domiciled in a UPCA contracting state is now exposed to UPC adjudication on the UK, Swiss, Turkish or Spanish part of any European patent at issue. And the validity defence will be governed by the substantive patent law of the State in question (the lex loci protectionis).
In practice, a Düsseldorf or Munich panel may have to decide novelty and inventive step under the UK Patents Act 1977, Swiss patent law, or Turkish patent law, in the same proceedings in which it applies the EPC and UPCA at home. That comparative workload is new. It is where teams using AI-assisted research platforms like Solve Intelligence have a real advantage over manual workflows.
The Düsseldorf Local Division did not wait long. On 28 January 2025, four weeks before BSH was handed down, it issued its decision in Fujifilm v Kodak (UPC_CFI_355/2023), concerning EP 3 594 009 B1.
The first headnote is now widely cited:
“If the defendant is domiciled in a Contracting Member State (here: Germany), the Unified Patent Court has jurisdiction to hear the infringement action in respect of the UK part of the patent in suit. This also applies if the defendant has filed a counterclaim for revocation in respect of the German part of the patent in suit.”
In practice, the Düsseldorf LD revoked the German part of EP 009. It confirmed jurisdiction over the UK designation because Kodak was domiciled in a UPCA contracting state. This is the first explicit UPC application of the long-arm principle to a non-UPC, non-EU State.
Counsel running parallel UK High Court actions now have to ask whether the UPC has already preempted the dispute. They also have to ask the harder question: can the UPC’s broad-construction default and its expansive view of implicit disclosure swing a UK validity ruling that the English court would not have reached on the same prior art?
Answering that requires cross-checking UK case law on novelty, inventive step and added matter against the UPC’s emerging jurisprudence. That is work AI-assisted research platforms like Solve Intelligence are built to compress.
Fujifilm is the most-cited example, but it is not isolated. A growing number of orders confront jurisdiction over non-UPC designations as a matter of routine.
In Mul-T-Lock v IMC Créations (Paris LD, 21 March 2025), the Mul-T-Lock entities filed a preliminary objection contesting UPC competence over the Spanish, UK and Swiss designations of EP 4 153 830. The Paris LD addressed all three. The Swiss-designation question raises the additional layer of the Lugano Convention, the parallel jurisdiction regime governing Switzerland, Norway and Iceland. That is territory the UPC will have to map out properly before long.
In the Sumi Agro v Syngenta saga, beginning June 2024 and continuing into 2026, jurisdiction over non-UPC designations and the timing of preliminary objections have been litigated repeatedly.
By our count of the UPC database through 4 May 2026, 14 decisions issued since BSH expressly engage jurisdiction over a non-UPC designation (UK, Swiss, Turkish or Spanish). 31 decisions cite BSH itself. The cross-border thread is no longer exceptional. It is becoming part of the standard fact pattern for any multi-jurisdiction infringement action. And each of those cases pulls a third-State validity question into the proceeding. Manual review of comparative validity standards at that scale is no longer realistic.
On 6 March 2026, the Court of Appeal in Dyson Technology v Dreame International (UPC_CoA_789/2025) issued two orders. The first, on the merits of a provisional-measures appeal, refined claim construction and the scope of injunctions under Article 62(1) UPCA. The second, issued the same day, stayed parts of the appeal and referred questions to the Court of Justice of the European Union.
It is the first CJEU referral in the UPC’s history.
The referral targets two outer-edge questions on long-arm jurisdiction:
The CoA ring-fenced the referral carefully. It did not stay the action against Dreame International with respect to the UPC Territory itself, nor against the Swedish and German Dreame entities. Provisional measures against the principal UPC-Territory infringing acts continue.
The CoA chose to refer rather than to extend BSH by analogy. That signals the UPC bench takes Article 71b Brussels Ia and the CJEU’s institutional prerogatives seriously, even where it would have been doctrinally tempting to push BSH further alone.
The Eurep “intermediary” question is the more interesting of the two. A substantive ruling on the role of EU-resident authorised representatives in patent injunction practice would likely resonate beyond Dyson, into other regulatory regimes where EU representatives are mandatory: medical devices, pharmaceuticals, and broader CE-marking compliance.
A CJEU ruling is unlikely before 2028. In the meantime the BSH baseline applies and the UPC continues to issue long-arm injunctions where the defendant is UPCA-domiciled. The signal to defendants is not “wait for clarity”. It is “litigate now under the existing baseline and prepare to revisit if Luxembourg moves the line.”
The practical consequences for European patent practice are concrete.
A single UPC filing can now plausibly produce relief across the UPC Territory plus the UK, plus Switzerland and Turkey where the defendant is UPCA-domiciled. The question of Spain remains pending the Dyson answers. The cost-per-jurisdiction calculus has shifted decisively in the UPC’s favour against parallel national filings.
The validity defence is now a multi-jurisdictional exercise. Because the UPC will apply lex loci protectionis on third-State validity, defendants must plead invalidity under UK, Swiss, Turkish or Spanish patent law (depending on which designations are in suit) in the same proceedings where they address UPC and EPC standards.
Central revocation in Paris or Munich becomes a more pressing counter-measure, because invalidating the European patent at source forecloses both home-jurisdiction injunctions and long-arm cross-border claims simultaneously. Preliminary objections on jurisdiction must be filed sharp and early; the Mul-T-Lock timeline shows the Local Divisions are willing to resolve them quickly.
The opt-out calculus has shifted. Patents opted out in 2023–24 were taken out of the UPC mainly to avoid central revocation. But an opted-out patent is also outside the UPC's new long-arm reach into the UK, Switzerland, Turkey and Spain. That is a benefit that did not exist when the original opt-out decisions were made. Portfolios opt out decisions made before BSH are worth a fresh look.
The pace and the breadth of UPC practice have both stepped up. New long-arm rulings appear most weeks, often buried inside orders captioned as provisional measures or claim construction rather than jurisdiction. At the same time, every long-arm action drags UK, Swiss, Turkish or Spanish validity standards into a single proceeding. That combination of high decision volume, scattered reasoning, multi-law substantive work is what AI-assisted research platforms like Solve Intelligence are built to compress.
The next two years will turn on three open questions. First, how far the CJEU is willing to push the BSH logic when the Dyson questions return. Second, whether the European Commission’s IPRED review, signalled by its January 2026 application study, layers an explicit proportionality test on top of the now-broader injunction power. Third, whether the UK Supreme Court overturns the Court of Appeal's Tesla v InterDigital & Avanci ruling and reshuffles the global FRAND landscape in ways that pull some cross-border disputes back toward London.
Through all three, the UPC will continue to be the courtroom most likely to surprise. After three years, the platform is no longer being judged on speed or predictability. Those promises are largely kept. It is being judged on its reach. On that measure, year three has been the most consequential year yet.
Solve Intelligence is purpose-built for the comparative, multi-jurisdiction work the UPC now demands. In particular, the third-State validity analysis that BSH has put on every long-arm defendant’s plate:
If your team is preparing UPC actions in 2026, book a demo to see how Solve Intelligence sits alongside your existing case-management workflow. The teams that get this infrastructure in place now will be the ones moving at the court’s speed when the Dyson answers come back.
Does BSH v Electrolux mean the UPC can grant injunctions covering the UK?
Yes, where the defendant is domiciled in a UPCA contracting state. Fujifilm v Kodak (Düsseldorf LD, 28 January 2025) was the first explicit application. The UPC cannot order any amendment of the UK register; validity determinations operate inter partes only.
What law governs validity when the UPC adjudicates a non-UPC designation?
The substantive patent law of the State of registration (lex loci protectionis). For the UK part of a European patent, the UK Patents Act 1977 governs; for Switzerland, the Swiss Patents Act; and so on. The UPC’s role is to apply that national law inter partes, without altering the foreign register. Defendants therefore need their validity defence pleaded under each relevant national standard, not only EPC/UPC standards.
What is the timeline for the Dyson v Dreame CJEU referral?
The referral was made on 6 March 2026. CJEU references in patent matters have historically taken 18 to 30 months. A ruling before mid-2028 would be unusually fast. The BSH baseline continues to govern UPC long-arm practice in the meantime.
How does Solve Intelligence keep current with new UPC decisions?
Solve Intelligence ingests UPC decisions from the official registry, structures them by court, division, parties, judges and citations, and surfaces relevant orders alongside the user’s existing matter context. A new long-arm order from a Local Division is searchable and citation-mapped within hours of publication.
Can Solve Intelligence help draft a UPC preliminary objection on jurisdiction, or a multi-law validity defence?
Yes to both. The drafting assistant produces first drafts of preliminary objections grounded in the relevant Brussels Ia provisions and the post-BSH UPC case-law, with citations to the LD and CoA orders most likely to control. For validity defences, Solve Intelligence generates parallel arguments under UPC/EPC and the relevant third-State patent laws (UK, Swiss, Spanish, Turkish), with the national case-law authorities surfaced in line. The output is structured for direct integration into a UPC CMS submission.
.webp)
