This article is part of the UPC series. It deep dives into the recent procedural order of the MIilan Central Division on addressing “divisional game” complaints before the Court.
As patent practitioners, we have all heard of the “divisional game”. Divisional applications are, in themselves, a legitimate and well-established feature of the patent system, allowing applicants to pursue distinct aspects of their disclosure in separate proceedings. The difficulty lies in striking the right balance between that legitimate function and its potential misuse: a patent owner may (i) file divisional patents that largely overlap in content, and so share the same weak points on validity, and then (ii) obstruct effective legal review of those weaknesses by withdrawing patents before the competent appeal instances can issue a decision. Used in this way, the mechanism can head off harmful precedents and ultimately shield the patent owner’s ability to enforce and defend the remaining divisional patents, which is why identifying the line between legitimate prosecution strategy and abuse of the system is not always straightforward.
The European Commission has taken notice of this practice and, in certain circumstances, has sanctioned it as a breach of Art. 102 TFEU (see, notably, the Commission’s decision in the well-known “Teva Copaxone” case of 31 October 2024). This remedy, however, only comes into play once the conduct has already caused damage to competitors, and takes the form of a public fine that does not directly compensate the competitors for the resulting delay in market entry. Meanwhile, the EPO has ultimately declined to introduce a fixed time limit for filing divisional applications, which therefore remain possible today provided that a parent application is still pending under Rule 36 EPC. A 24-month deadline[1] had in fact been in force between 2010 and 2014, but the EPO abolished it as of 1 April 2014.
Against this background, a procedural order issued by the Milan Central Division on Thursday, 22 October[2], is attracting attention: it may well mark the first milestone of a new strategy for claimants seeking to clear their way ahead of market launch.
The claimant’s requests were quite bold. It asked the Court to: (i) declare invalid, for the same lack of inventive step invalidating the patent at issue, any claim of any patent entitled to the same priority date as an application within its patent family, insofar as that claim covers the same subject matter (“Request 6”); and (ii) order the patent owner to refrain from withdrawing the patent before a final decision on its validity has been issued (“Request 7”). At the oral hearing, the claimant openly stated that both requests were aimed at countering a divisional game played by the patent owner. The first request appears designed to have the Court take a position on a question that divides practitioners: whether an invalid divisional with overlapping subject matter also affects the validity of that same subject matter in the parent. The second is aimed at preventing precisely the conduct addressed by the Commission in Teva Copaxone, namely withdrawal of the divisional before a decision has been issued, thereby protecting the entire patent family.
The defendant raised preliminary objections against such requests by substantiating equally reasonable grounds, inter alia, that that the claimant cannot request open-ended declarations concerning future patents or claims, which may not even fall within the UPC jurisdiction, and that neither the UPCA nor the RoPs give the Court any power to interfere with the EPO procedure to the point of impeding the patent owner to withdraw a text from approval, which is possible under Art. 113 (2) EPC.
After acknowledging and extensively describing the “divisional game” problem in the system, the Court rendered its procedural order, finding that it has jurisdiction to evaluate the claimant’s request. The principles set forth by the Court can be summarized as follows:
- The Court’s “inherent jurisdiction” is the key. Relevant to both Requests 6 and 7, the Court held that Art. 32(1) UPCA, as a provision of an international treaty, must be interpreted autonomously under Art. 31 of the Vienna Convention on the Law on Treaties, hence in good faith and in light of the object and purpose of the agreement itself. On that basis, the Court reasoned that a judicial body entrusted with exclusive jurisdiction over revocation actions, DNIs and provisional measures must also possess the inherent procedural authority necessary to protect the integrity and effectiveness of the proceedings duly brought before it. The Court grounded this conclusion in a joint reading of the recitals to the Unitary Patent Package, the UPCA preamble’s commitment to effective enforcement and legal certainty, and, crucially, Art. 47 of the Charter of Fundamental Rights, which guarantees the right to an effective remedy before a tribunal; read together with Art. 17 of the Charter on the protection of property including intellectual property, these provisions were held to require that judicial protection within the unitary patent system be effective in practice, since the system’s purpose would be impaired if the Court were powerless against procedural strategies depriving its validity jurisdiction of practical effect (such as a potential “divisional game” strategy). The Court further supported this approach by reference to the Munich Local Division’s anti-anti-suit injunction practice in Huawei v Netgear as an example of the Court protecting the effectiveness of its own proceedings against external interference even absent an express literal wording in Art. 32 UPCA . On such basis, Request 6 falls within the Court’s jurisdiction to the extent that it seeks to preserve effectiveness of the pending revocation proceedings should the patent owner withdraw the patent at dispute while at the same time pursuing equivalent subject-matter in the same family, through divisional applications. The Court also recalled the Omnia v- Sidel very recent case-law, to argue that jurisdiction can be established even with respect to not-yet granted patents.
- The Court’s power to restrain withdrawal of the patent text under 113(2) EPC is an exercise of inherent jurisdiction. The Court approached Request 7 through the same inherent-jurisdiction lens as Request 6, but anchored it more precisely. It located the power not in Art. 32(1)(c) UPCA (the general head of competence for provisional and protective measures, ordinarily exercised by the Local or Regional Divisions in relation to infringement), but in Art. 32(1)(b) and (d) UPCA, which confer on the Court jurisdiction over actions for declarations of non-infringement and revocation actions respectively: the effectiveness of those proceedings, the Court held, may require measures capable of preserving their subject matter and preventing procedural frustration, and such powers are a direct emanation of, and inherent in, the effective exercise of those heads of jurisdiction. On this basis, the Court regarded its jurisdiction to assess Request 7 as, in any event, inherent in its power to ensure that a revocation action properly before it remains effective.
- The Court might refuse to dispose of an action under Rule 360 RoP, as the claimant might still have an interest in the action when a divisional game is played. Here the Court made a further, interesting, step, besides the claimant’s requests. It observed that under Rule 360 RoP the Court may dispose of an action only where it “finds” that the action has become devoid of purpose, and that this finding is for the Court to make, not an automatic consequence of, e.g. the withdrawal of the patent in suit. This means that, where an application to dispose of the action under Rule 360 RoP is based on circumstances arising from a strategic use of divisional filings, the Court may reject that application and proceed to determine the merits. In other words: if ruling on the validity of the subject matter at issue remains useful precisely because that subject matter could otherwise be redeployed through other divisional patents, the action may not be devoid of purpose.
It is of paramount importance to note that this is not a decision on the merits: at this stage, the Court has not issued any order against the patent proprietor. Indeed, it stated that, based on the evidence provided by the claimant (namely, the applicant’s reservation of the right to file divisional applications in its reply to the EPO’s Communication pursuant to Rules 69 and 70a(1) EPC), it could not be concluded that the patent holder had engaged in a ‘divisional game’. This procedural order is nonetheless unprecedented, and its fate on any appeal will certainly be followed closely by patent lawyers and attorneys alike.
[1] 24-months from the date of a first EPO action on the merits (i.e. action raising patentability objections or confirming patent eligibility of the claims) or 24-months from date on which a non-uniformity objection (i.e. objection that the set of claims includes more than one invention) was raised for the first time.
[2] UPC_CFI_2221/2026, Procedural Order (on a Preliminary Objection under R. 19 RoP), Court of First Instance of the Unified Patent Court, Central Division (seat in Milan), delivered on 22 October 2026.
