UPC Unfiltered, by Willem Hoyng – UPC decisions week 04, 2026

15 January 2026 (late published)

Local Division Brussels, Labs v GC Aesthetics

UPC_CFI_1357/2025

Preliminary objection

Facts

1. Labs started infringement proceedings against 13 companies in various (also non-UPC) countries.
2. Labs argues that the Local Division has (international) jurisdiction, because one of the defendants is a Belgian company and there is a commercial relationship between all defendants.
3. GE Aesthetics lodges a preliminary objection with respect to the Irish, Spanish, Norwegian, Swiss, and UK designations of the invoked patent, because the Belgian defendant has no activities outside Belgium (let alone outside the UPC area). They refer to the decision of the Court of Justice of the European Union (“CJEU”) in Roche v Primus (case C-359/03).
4. Labs argues the objection relates to the scope of the ordered measures, and not to the jurisdiction itself. Moreover, the Belgian defendant “engages in coordinated commercial activities, including supply-side act by UPC-domiciled entities”.  

The Court

1. The defendant (Labs) does not object to the international jurisdiction and competence of the Brussels Division with respect to the defendants established in non-UPC countries, but only to the jurisdiction with respect to alleged infringements in non-UPC countries.
2. As no invalidity arguments have been made, the Court sees no problem in dealing with the infringement.
3. The Court dismisses the preliminary objection.

Comment

1. The Brussels Division is right: if you do not argue that the Court has no international jurisdiction and/or internal competence with respect to defendants then such jurisdiction is a given. The Court refused to help the defendants in reading something in the requests that was not there. This once again shows that you should be an experienced litigator if you litigate before the UPC. In this case it could have been wise to ask advice from a Dutch litigator; they have experience with cross-border issues since the nineties (both cited CJEU cases Roche v Primus and Solvay v Honeywell come from Dutch courts). The questions referred to the CJEU in the latter case were formulated by judge Kalden who is now in the Court of Appeal of the UPC.
2. It should be noted that the more recent CJEU case BSH Hausgeräte v Electrolux was about jurisdiction based on Art. 4 of the Brussels I bis Regulation (“BR”), and not Art. 8(1) BR.
3. In Roche v Primus, the CJEU held that there cannot be a situation of conflicting judgments (and thus applicability of Art. 8 (1) BR) if each defendant is only infringing in his own country. So, indeed, if the Belgian defendant would only infringe in the UPC and all non-UPC defendants would not infringe in the UPC, there would be no international jurisdiction based on Art. 8 (1) BR against these non-UPC-based defendants, if one applies Roche v Primus. However, if the claimant would state that such (cross-border) infringements take place, then in general, jurisdiction is given (except if it is obvious that the statement is untrue).
4. First of all, the Roche v Primus situation (which was created by the question phrased by the Dutch Supreme Court which had stated as facts that each defendant only infringed in their own country) shall in practice not happen very often in these modern internet times, where often products are offered on websites accessible in many countries. Secondly, it is very much the question whether or not Roche v Primus is still good law. The decision has been heavily criticized. It is of course very strange to say that decisions are not conflicting if one court in the EU – with respect to the same patent and the same infringing product, and applying the same law (Art. 69 EPC) – comes to a different decision than another EU court. In my opinion, later CJEU cases such as Painer (a copyright case) indicate that Roche v Primus is not good law anymore. However, that remains to be confirmed by the CJEU.
5. The decision in BSH Hausgeräte v Electrolux is also of importance in an Art. 8 (1) BR case when dealing with a validity challenge. In an obiter dictum, the Brussels Division in fact stated that the invalidity of the European patent in the various countries was not challenged, and that therefore the Division can proceed with the infringement. If challenged, the Division would have the option to e.g. suspend the infringement action or (with respect to non-UPC countries) deal with the invalidity defence.
6. The fact that the Brussels Division is not competent to deal with the validity of European patents granted for non-UPC, EU and Lugano countries does in my opinion not have to be raised already in a preliminary objection against an infringement claim, but can be done in the statement of defence to the infringement claim. Often a revocation action with respect to the patent will be filed, which seems wise if you want to argue that the European patent is also not valid for non-UPC countries! 

16 January 2026 (late published)

Local Division The Hague, BTL v Lexter

UPC_CFI_1048/2025

Settlement

Facts

The claimant asked for withdrawal of its claim and the defendant consented. The parties did not want a cost decision.

The Court

1. Compliments for the German company BTL which did not file in a busy German Division, but against the Spanish company Lexter in the The Hague Division!
2. Apparently the case was settled during the oral procedure because no reimbursement of fees was requested (although that may still be done!).

16 January 2026 (late published)

Local Division Paris, IMC v Mul-T-Lock  

UPC_CFI_702/2024 ; UPC_ CFI_369/2025

Infringement / revocation

Facts

1. The infringement action started on 20 November 2024, on the basis of the Unitary Patent in the UPC and on the basis of the European patent in Switzerland, Spain, and the UK.
2. Defendant filed a counterclaim for revocation with respect to the Unitary Patent.
3. The claimant IMC replied with four auxiliary requests. The fourth auxiliary request was of general nature, requesting the Court to amend the patent in a way that a valid patent remains. 

The Court

1. After describing the contents of the patent, the Court cites the claims as unconditionally amended by the patentee.
2. The Court defines the skilled person, referring to the decisions of the Court of Appeal in Amgen v Sanofi and Meril v Edwards.
3. The Court interprets the claims and rejects the limiting interpretation argued by the defendant, relying on the drawings which according to the Court show different possible realizations of the invention with other advantages of the invention.
4. The Court accepted during the oral hearing two new arguments with respect to an already disputed claim element by the defendant, which arguments were rejected.
5. With respect to the revocation, the Court stated that it is the claimant in the revocation action which has to prove the common general knowledge (referring to a decision of the Paris Central Division in NJOY v VMR (UPC_CFI_310/2023).
6. The Court rules that the Unitary Patent is valid in the amended version. In addition, it ruled the patent is infringed.
7. The fact that the Unitary Patent was maintained in an amended version raises serious questions about the validity of the European patent in Switzerland, and the Court is therefore not capable to rule on the infringement in Switzerland.
8. The Court applied the Court of Appeal’s decision in Belkin v Philips stating that what constitutes and infringing offer has to be interpreted in an economic sense and not in a strictly legal sense.
9. The amended version is valid as of the day of the grant of the European patent, and not as defendant argued from the day of the amendment.

Comment

1. A well-reasoned decision, in my opinion, and delivered within 14 months!
2. I think that also in light of the CJEU’s decision in BSH Hausgeräte v Electrolux, a more practical solution for (in this case) Switzerland would have been to hold that the Swiss part of the European patent is valid and that the validity has not been challenged in a Swiss court by defendant. Moreover, it is clear that (unless defendant would convince the Court of the opposite) also in Switzerland after invalidity proceedings the patent would survive in the amended form. So it seems justified to grant the injunction for Switzerland as well.
3. It is not clear what happens with respect to the UK and Spain. Apparently the claim with respect to these countries was withdrawn.
4. One can even add to the above the condition that the claimant has to file central limitation proceedings in the EPO or undertake steps in Switzerland within two months to amend the European patent, or grant the injunction on the condition that the injunction becomes effective as soon as the (Swiss part of the) European patent has been amended (in the same way as the unitary patent).
5. However, I do not think that these further conditions are necessary.
6. With respect to non-EU or Lugano countries such as the UK, the Court can establish the validity inter partes. In such a case, the Court can also hold the patent valid in amended form inter partes.
7. However, as I have said before, I do not think that the UPC should grant injunctions against local UK companies for reasons of comity (or proportionality), except where such local company is the subsidiary of a parent company established in the UPC territory. With respect to EU but non-UPC defendants (such as Spanish or Polish defemdants) the same problem does not arise because competence of the Court is based on the Brussels I bis Regulation.
8. Finally, I do not think, as the claimant asked in this case, that an auxiliary request can (only) be a request to the Court to amend the patent in such a way that it is valid! On the other hand, if an auxiliary request could lead to a valid patent after a small change, I have no problem with the Court suggesting so during the oral argument (assuming defendant gets sufficient time to react). However, it is of course the task of a representative to draft a limited number of good auxiliary requests without turning the UPC into the EPO where patent attorneys can file an (almost) unlimited number of auxiliary requests. 

16 January 2026 (late published)

Court of Appeal, Applicant (X) v Amycel

UPC_CoA_935/2025

Suspensive effect / default judgment

Facts

1. The Local Division of The Hague granted a PI against X after inter partes proceedings on 31 July 2024.
2. On 30 August 2024, Amycel started proceedings on the merits against X.
3. The Local Division of The Hague considered the Statement of Claim validly served via an alternative method of service (R. 275(3) RoP).
4. By order of 1 April 2025, the Local Division of The Hague ordered a decision by default to be taken because X’s representative had not timely filed a defence.
5. The decision by default was issued on 21 October 2025.
6. X filed an appeal against certain parts of the default judgment and asked for suspensive effect, basically arguing that it would be impossible to undo the decision as the message it was ordered to put on the internet would always stay on the internet, and lack of urgency.

The Court of Appeal

1. The Court of Appeal dismisses the request for suspensive effect.
2. X failed to establish that the decision is manifestly erroneous. The consequences of the order do not become devoid of purpose even if the consequences of these orders (such as the sending of registered letters and putting messages on the internet) are not fully reversible. 

Comment

1. A decision by the full panel of the first chamber of the Court of Appeal which again shows the very exceptional nature of suspensive effect.
2. One can even see that irreversible consequences as such are not good enough if you do not show that the first instance decision is manifestly wrong.
3. Indeed, one cannot say that an appeal becomes devoid of purpose. If the appeal is successful, the Court of Appeal can order to send a correction letter to clients and also place a correction on the website. So irreversibility cannot be taken seriously.
4. The representative of X should have known better. It would have been much better for X to (at least after the PI proceedings) seek a settlement and avoid Amycel to make more costs in proceedings on the merits against which X did not want to defend himself.
5. As I understand this decision, you can appeal a default decision or you can ask the Division which granted the decision to set the decision aside. I find this somewhat strange. It seems to me that it is more logical that somebody first requests to set the decision aside. If it is correct that on appeal you cannot add new facts, then it seems to me that the appeal is entirely fought on the basis of the facts as put forward in first instance by the claimant. Very strange!

19 January 2026

Local Division Mannheim, Huawei v HMD

UPC_CFI_481/2025

Confidentiality and FRAND 

Facts and order
In the past, the Court already ordered confidentiality with respect to certain parts of the FRAND negotiations between the parties. 

The JR proposed (and the parties accepted) a general confidentiality order also covering future exchanges with respect to the FRAND negotiations. 
The JR indicates that a request for a confidentiality order with respect to other information remains possible. 

Comment

A practical solution saving the very busy Local Division of Mannheim some work. However, it would have been much better if this case between a Chinese company and a Finnish company would have been filed elsewhere, which is perfectly possible as all Divisions accept the English language!

23 January 2026 (not yet published)

Local Division Paris, Guardant v Sophia

UPC_CFI_808/2025 

Condition for PI

Facts

1. Guardant starts preliminary injunction (“PI”) proceedings on the basis of four European patents against four Sophia companies (French, Swiss, Italian and German) for the UPC and various non-UPC countries.
2. The oral hearing took place on 12 December 2025.
3. Guardant withdrew its claims with respect to one European patent.

The Court

1. With respect to urgency (“any unreasonable delay in seeking provisional measures” under R. 211.4 RoP) the Court cited the Court of Appeal’s decision in Mammut v Ortovox of 25 September 2024.
2. Information about the infringement from before 27 May 2025 is either too general or relates to activities outside Europe. The manual was online as early as October 2024, but Guardant cannot be expected to monitor the internet for all competing products without having been alerted by Sophia’s activities in Europe.
3. The Court accepts urgency as a three months delay to prepare an action invoking several patents and complex technology is reasonable.
4. The Court deals with each of the patents separately. It establishes the skilled person and interprets the claims applying the Court of Appeal’s decision in Nanostring v 10x Genomics (UPC_CoA_335/2023).
5. A method which is mentioned in the description as other method, but is not mentioned in the claim, does not fall under the scope of protection.
6. With respect to added matter, the Court refers to the principles formulated by the Court of Appeal in the decision in expert klein v Seoul Viosys (UPC_CoA_764/2024).
7. The Court disagrees with the Opposition Division. It is insufficient to find all parts of the claim scattered over the whole PCT application. The question is if the claimed invention can be found in the document.
8. The Court concludes with respect to EP 3 591 073 that it is more likely than not that the patent is invalid and not more likely than not that it is infringed.
9. With respect to EP 3 766 986, basically for the same reason as cited under 7, the Court comes to the conclusion that the patent is more likely than not invalid for extension of subject matter.
10. With respect to the third patent, EP 3 443 066, the Court concludes that Guardant has not (sufficiently) proven the way Sophia’s software processes the information. The press release on which Guardant relies, gives insufficient details as to what is done and what information is stored in a database.
11. The Court rejects the application for a PI. The Court reduces the requested (interim) costs: € 400.000.

Comment

1. The Division has a very realistic view on urgency. We have seen less realistic applications in for instance Brussels and Lisbon.
2. I wonder why the claimant chose to file PI proceedings on four patents in rather complicated proceedings without first properly establishing the facts (see for example above under 10).
3. I fully agree with the Court that there is extension of subject matter if the invention which is claimed is not disclosed in the application (which may be several hundred pages). The assessment is not limited to locating individual claim elements, potentially dispersed throughout the application, but also requires determining whether the skilled person, reading the application as a whole, would understand it to disclose an invention consisting of the specific combination of those elements.  

– All comments above are Prof. Hoyng‘s personal opinions –

For all decisions of the UPC’s Court of Appeal from the start in 2023 until 1 January 2025, Willem Hoyng’s more comprehensive summaries and comments are bundled in the book “Case law – Court of Appeal of the UPC 2023-2024”.