UPC draws sharper lines on confidentiality clubs and security for costs
On 18 June 2026, the UPC released its latest monthly statistics. The more important development for practitioners, however, was not the headline caseload but the procedural message coming through a cluster of recent orders. The Court is drawing more precise boundaries around confidentiality review, front-loaded pleading discipline and security for costs. For businesses handling SEP, FRAND, infringement-plus-revocation or multi-country European patent disputes, these are no longer background issues. They affect defence timing, litigation budgeting and who inside the company can be involved, and when.
Two signals stand out. First, in disputes that turn on comparable licence agreements, the Court is increasingly willing to let an initial review be conducted by external counsel and independent experts before deciding whether any wider access is justified. Second, security for costs analysis is moving back to recoverability in practical terms rather than abstract arguments about nationality or place of incorporation. Pointing to a European bundle patent portfolio is not, by itself, the same thing as offering a reliable and executable cushion for a future costs order.
Confidentiality clubs are becoming staged rather than all-or-nothing
The UPC has had confidentiality tools from the outset, but recent orders make the operating logic easier to see. The Court is increasingly separating two questions that companies often try to collapse into one. The first question is which documents are truly relevant, which licences are actually comparable and which material needs to enter the merits debate at all. The second is who should be allowed to see that material once those thresholds are cleared. That sequencing matters. In the first phase, external representatives and independent experts are becoming the natural first readers, while in-house technical, business or licensing personnel are less likely to be admitted automatically.
That should not be misunderstood as a permanent rejection of party employees. The Court of Appeal has already made clear that, as a general rule, at least some form of employee access may be necessary to preserve the right to an effective remedy and proper conduct of the case. What is changing in practice is narrower and more important: where comparable licence agreements have to be screened first, the Court is willing to accept an external-eyes-only phase before deciding whether internal personnel should be added later, and on what conditions. That is a meaningful shift for companies that still assume their commercial teams will be inside the room from the start.
This changes FRAND disclosure strategy and raises the cost of delay
In SEP and FRAND disputes, licence disclosure has always been difficult. The sensitivity does not lie only in the contract itself. Comparable licences usually carry third-party commercial terms, pricing logic and negotiating history that none of the parties can treat casually. A staged review by external lawyers and independent experts gives the Court a way to filter out agreements that are ultimately irrelevant or not sufficiently comparable before they are exposed more widely. From a procedural perspective, that is a cleaner solution. From a commercial perspective, it is also easier to defend vis-à-vis third parties.
Set beside another recent procedural ruling from Paris, the broader message is hard to miss. The UPC remains a front-loaded court. Parties are still expected to bring core non-infringement arguments and evidence early, not after the written phase has largely taken shape. Put together, these developments mean that defendants cannot simply wait for the confidentiality architecture to settle and only then decide on their technical line of defence. External experts, FRAND economists, licensing specialists and internal legal teams need to be aligned much earlier. The old instinct to file first and refine later fits this forum badly.
Security for costs turns on real recoverability, not on foreign status alone
The Court’s recent approach to security for costs is also becoming more concrete. The UPC is not treating a claimant’s status as a U.S., Chinese or otherwise non-European entity as an automatic problem in itself. Foreign domicile alone is not enough. What matters is whether there is a legitimate and real concern that a future costs order would be difficult to recover in practice. That inquiry is highly fact-specific. It looks to funding structure, liquidity, asset quality, enforceability and whether the claimant’s apparent assets are already heavily encumbered.
That is why arguments based simply on the existence of a European patent portfolio are losing force in this context. A patent portfolio may well be an asset in accounting or valuation terms. But for security purposes, the Court is asking a different question: does it function as a realistic backstop if the claimant loses and costs are awarded? If patents, royalties, licence income or disposal proceeds are already pledged to a third party, the entity may still be treated as economically asset-light or close to assetless. For non-European claimants, especially newly formed SPVs or patent assertion vehicles with complex financing structures, that is no longer a theoretical risk. It is a budgeting issue that needs to be faced before proceedings start.
The practical response now is to move three tasks forward
First, any case that may involve third-party licence agreements should be prepared with a layered disclosure plan before the claim is filed or the defence is finalised. Which agreements are likely to be relied on as comparables, which fields are most sensitive, what can be summarised by experts in a first phase, and which internal personnel could later be proposed for access all need to be mapped in advance. Second, parties planning to sue at the UPC, particularly claimants based outside Europe, should assess early what form of security they could provide if required, whether that means an EU-bank guarantee, a cash deposit or clearly documented financial support.
Third, teams need to treat the UPC’s procedural timetable as genuinely integrated. Confidentiality disputes, revocation counterclaims, non-infringement arguments and licensing analysis are not separate workstreams that can be handled one after another. They now interact from the start. That is the real significance of this June line of decisions. The Court has not made proceedings looser. It has made the boundaries clearer: who may see what first, which arguments must be ready earlier, and which claimants may be asked to put a costs safety net on the table at the outset.
