The European Union seems to be revisiting an old debate about injunctions and injunctive relief in patent disputes. Although the debate focuses on core questions about remedies available to patent owners, it now takes place in an environment of increasingly sophisticated technology markets and complex value chains. In the past years, there have been several attempts to weaken patents and their enforcement in the EU – most notably through the 2024 proposal to reform the system of Standard Essential Patents (SEPs) in the ICT sector. Such efforts, however, faced significant opposition. The experience of the ICT sectors demonstrates that robust patent protection and effective enforcement are key to innovation, competition, and economic growth. Indeed, the patent licensing ecosystem in these industries has proven remarkably capable of producing both upstream technology development and downstream use.
Is this time different? Some would have us think so. In January 2026, the European Commission published a follow-up study on the application of the Intellectual Property Rights Enforcement Directive (IPRED), a 2004 directive that aimed to harmonise the enforcement of intellectual property (IP) rights across member states of the European Union. The study examined several aspects of the Directive’s operation, including the reliance on proportionality in patent injunction cases, Patent Assertion Entities (PAEs), dynamic blocking injunctions, and the costs of destroying infringing goods. The renewed debate over IPRED reform has increasingly become centred on one question: the courts’ approach in applying the proportionality principle when deciding whether to grant injunctions in patent infringement cases.
There are several ways of looking at this issue – some useful, others not. For some, the question arises from a perceived conflict between Article 3(2) of IPRED, which requires enforcement measures to be effective, proportionate and dissuasive, and Article 11, which makes injunctions available as a remedy against infringement. In the UPC system, the same issue is reflected in Article 63 of the Agreement on a Unified Patent Court (UPCA), which provides that the Court may grant an injunction after a finding of patent infringement. In a way, this settles the principal legal question. It is not about whether proportionality exists in the EU framework, because it does, but how it is and should be applied in patent cases.
Still, the debate lingers and largely reflects differing views among those who seek to limit the availability of injunctions through an expanded application of the proportionality principle and those who believe the existing framework is functioning effectively, with courts continuing to develop jurisprudence that provides increasing clarity and predictability regarding the principle’s application. This divide was also reflected in the recent European Parliament’s JURI workshop of 3 June 2026 on the evaluation of IPRED, where the Committee on Legal Affairs examined the Directive’s legal assessment, effectiveness and possible reform.
The debate, however, is not new. It sits within a longer, two-decade effort to reconcile two objectives that may sometimes pull in different directions: the effective enforcement of IP rights across the internal market, and sufficient flexibility for courts to prevent disproportionate outcomes in individual cases. The Commission’s 2017 guidance on IPRED was itself an attempt to manage that tension, including by clarifying the role of proportionality and safeguards against abuse in the application of enforcement measures. Yet the fact that proportionality remains a point of discussion does not mean that IPRED is defective or that judicial practice is materially divergent. Rather, it reflects the ordinary operation of proportionality as a general legal principle: courts apply it in light of the specific circumstances of each case, balancing the need for effective enforcement against other relevant interests where appropriate. Indeed, the available evidence does not demonstrate material divergence; if anything, it suggests that proportionality is already considered within the existing framework in a restrained and largely harmonised manner.
Importantly, patent disputes rarely present courts with a simple choice between enforcement and non-enforcement. Patent enforcement cases usually end up in court because they are very complex. And courts do have flexibility to balance competing objectives. For instance, the UPC’s decision in Meril Life Sciences v Edwards Lifesciences, concerning prosthetic heart-valve technology, illustrates how injunctive relief needs to take account of third-party and patient interests. The point is not that such cases should become the model for every patent dispute, but that they show why proportionality requires context-sensitive reasoning – not legislative restriction.
Proportionality in Practice
The principle of proportionality gives little, if any, basis for legislative reform, given that the principle is already built into IPRED’s remedial architecture, and adopted across member states, as the IPRED follow-up study itself recognises. As noted above, the Directive does not create an automatic or unconditional right to an injunction, but neither does it support a policy presumption against injunctive relief. Its basic compromise is clear: IP rights should receive strong and equivalent protection across the internal market, subject to safeguards against disproportionate or abusive enforcement.
That compromise is reinforced elsewhere in the Directive. Recital 17 states that remedies should be determined case by case, taking account of the specific characteristics of the case and the specific features of the IP right at issue, including, where appropriate, the intentional or unintentional character of the infringement. Article 12 also allows Member States, in appropriate cases, to provide for pecuniary compensation instead of applying corrective measures or injunctions where the infringer acted unintentionally and without negligence, where execution of the measure would cause disproportionate harm, and where financial compensation would be reasonably satisfactory.
This provision is narrow and does not displace injunctions as a central remedy. It confirms that IPRED was not drafted on the assumption that enforcement must be blind to proportionality.
This is also why proportionality should not be understood simply as a reason to deny injunctions. Germany’s amended § 139 Patent Act reflects a similar logic. The 2021 reform introduced an express proportionality limitation where enforcement would cause disproportionate hardship to the infringer or third parties not justified by the exclusive right. But it also provides that, where the injunction claim is excluded on that basis, the right holder may claim reasonable monetary compensation, while the ordinary damages claim remains unaffected. The point is therefore not to weaken injunctions as such, but to ensure that proportionality operates without turning valid patent rights into merely compensatory claims.
That matters because IPRED was adopted to ensure a high, equivalent and homogeneous level of protection for IP rights across the internal market. It was not designed to make infringement easier, or to weaken remedies once infringement has been established. Its purpose was to address disparities in national enforcement systems that undermined the internal market and reduced confidence in the protection of these rights. To the extent that differences remain across Member States, they appear to reflect limited variations in implementation or judicial articulation, not material divergence in the application of proportionality or any structural defect in IPRED itself.
Many of the arguments advanced in support of revising IPRED are based on a misreading or misinterpretation of the available evidence. And here it gets interesting: unfortunately, the Commission’s follow-up study illustrates this problem. That study draws on a mixed evidence base, including case-law review and stakeholder evidence, which can identify perceptions of uneven practice but cannot by itself establish that IPRED produces automatic injunctions across the EU. The headline figures that are often cited in this debate – that permanent injunctions were granted in 94 per cent of patent cases and that proportionality was expressly assessed in only 0.6 per cent – are misleading. They should not be read as showing that injunctions are granted in 94 per cent of all patent disputes.
In the UK, for example, only 4 of the 14 patents considered at first-instance infringement trials in 2025 were found valid and infringed. The corresponding figures were 4 of 21 in 2024, 6 of 22 in 2023, 7 of 27 in 2022, and 10 of 32 in 2021. The emerging UPC data points in the same direction: patent owners prevailed in 8 of 18 infringement cases in 2024 and on 20 of 55 patents in 2025. These figures suggest that injunctions are not being granted automatically across patent disputes; rather, an injunction remains contingent on the claimant first succeeding on validity and infringement. In Munich, for instance, 187 patent infringement claims were withdrawn before judgment in 2025, while only 80 proceeded to judgment.
A dataset focused on decided cases may therefore disproportionately capture claims that survived earlier litigation filters, while omitting disputes that were withdrawn, settled, dismissed, or otherwise resolved before injunctive relief became relevant. The relevant sample in the study appears to concern cases reaching the remedy stage, not patent disputes as a whole. In that context, a high injunction rate mainly reflects the fact that validity and infringement have already been established in the remaining cases. The figure therefore tells us less about automatic injunctions than about the composition of the cases being measured, making it unsafe to infer from judgment outcomes alone that proportionality is not being meaningfully considered.
The same critique applies especially to the 0.6 per cent figure. An express reference to “proportionality” in the court’s decision is not the only way a court may assess whether relief is appropriate. Courts address proportionality-relevant considerations through other concepts, including hardship, balance of interests, abuse of rights, public interest, third-party effects, FRAND conduct, stays of enforcement, or the availability of alternative remedies. A low rate of express citation is therefore consistent with more than one explanation: courts may be dealing with the issue through established domestic legal vocabulary or encountering few cases in which an explicit and serious proportionality defence is actually pleaded and evidenced.
Nor is a full proportionality analysis appropriate in every patent case. Where infringement is established and no serious proportionality objection is raised, an injunction may be the ordinary and proportionate remedy. The relevant question is therefore narrower: how courts reason in the subset of cases where the facts genuinely raise proportionality concerns? To answer that, the evidence would need to show not only how often proportionality concepts are addressed, but how often proportionality arguments were pleaded, what facts supported them, how courts addressed those arguments, and whether the remedy was tailored, stayed, refused, or replaced by compensation. On that question, the current evidence supports more careful data collection and closer attention to the arguments actually raised by the parties; it does not support a general legislative presumption against injunctions.
The reform debate should therefore distinguish between two different claims. One argument advanced by proponents of reform is that courts should engage more expressly with proportionality when granting injunctive relief. Yet the available evidence does not demonstrate inconsistent or inadequate judicial application of the principle. The low number of cases in which proportionality is expressly assessed is better understood as reflecting the limited circumstances in which parties raise a serious proportionality objection, rather than as evidence that courts are unable or unwilling to apply the principle properly. At most, this supports continued attention to how proportionality objections are pleaded and addressed in practice, a task that the developing case law of the UPC is well placed to perform. That is different from the more far-reaching claim that injunctions are too readily available as a matter of EU law and should therefore be legislatively restricted. The available evidence does not support that claim.
European Patent Enforcement is Already Evolving
Developments in national patent law, SEP case law, and the UPC point in the same direction: European patent enforcement is continuing to evolve in its application of the principle of proportionality. The German amendment, as discussed earlier, responded in part to concerns that injunctions in complex-product cases could create hold-up risks if applied too mechanically. However, the German experience does not show that proportionality was absent from European patent enforcement, nor does it establish a systemic problem of inconsistent judicial application. In the SEP context, courts across Europe have generally approached injunctions through the Huawei v ZTE framework, focusing on the parties’ conduct during FRAND negotiations. One may debate whether that framework places too much emphasis on negotiating behaviour, but it is difficult to characterise it as evidence of inconsistent proportionality analysis. The UPC further reinforces this point: as a single court exercising jurisdiction across participating Member States, it is institutionally designed to promote a more coherent and uniform body of patent jurisprudence, thereby reducing concerns about divergent interpretations of proportionality in patent disputes.
The development of the UPC further undermines this case for legislative intervention. The UPC system only began operating in June 2023, and its approach to remedies is still developing. Proportionality is already part of the legal framework within which it operates. Article 42 requires the Court to deal with litigation proportionately and to ensure that procedures and remedies are used in a fair and equitable manner and do not distort competition (along with Article 63 UPCA). The fact that the UPC’s case law on remedies is still developing is not evidence that the system is broken; it reflects the ordinary development of jurisprudence within a court specifically designed to promote coherence and uniformity across participating Member States. This is precisely what some advocates emphasised during the June 2026 workshop: the need to allow the UPC’s jurisprudence to develop further before considering amendments to the existing IPRED framework.
National courts confirm the same trajectory. In Germany, the Federal Court of Justice’s Heat Exchanger decision shows that proportionality considerations were not absent from the pre-reform patent enforcement framework: even before the statutory amendment, the Court accepted in principle that injunctive relief could be limited in exceptional cases where immediate enforcement would cause disproportionate hardship contrary to good faith, although it refused to grant such relief on the facts.
And in France the Paris Court of Appeal’s Peugeot v Piaggio judgment is better understood as a cautionary example than as support for reform. While the court refused a permanent injunction on proportionality grounds, relying on factors such as the imminent expiry of the patent, the limited number of infringing products, and declining market share, that reasoning also illustrates the danger of expanding proportionality too far. If damages are treated as an adequate substitute for injunctive relief in ordinary infringement cases, the patent right risks being reduced from an exclusionary right to a merely compensatory claim. The Peugeot case should therefore be treated as an exceptional, fact-specific decision, not as evidence that IPRED requires reform or that permanent injunctions should become more readily replaceable by damages. Taken together, these cases show that proportionality can be addressed through existing legal mechanisms and fact-sensitive judicial analysis, without the need for legislative intervention.
SEP litigation is perhaps the context in which we have observed the strongest convergence in the courts’ analysis and practice. The framework established in Huawei v ZTE places structured legal constraints, through competition law, on the pursuit of injunctions by SEP holders that have given a FRAND commitment: before seeking an injunction, the SEP holder must alert the alleged infringer and, where the infringer expresses willingness to take a licence, make a specific written FRAND offer; the infringer must respond diligently, in good faith and without delaying tactics. Injunctions in SEP cases are therefore not automatic but subject to a competition-law framework that scrutinises the conduct of both parties.
Over the years, courts have provided important guidance on how to apply the Huawei v ZTE framework in individual cases. In Nokia v Oppo and Conversant v Huawei/ZTE, courts emphasised that implementers must engage in genuine and target-oriented negotiations, rejecting attempts to delay or avoid concluding a licence. In Unwired Planet v Huawei, the UK courts confirmed that Huawei provides a “safe harbour”, whereby compliance reduces the risk of abuse without imposing rigid or formalistic requirements. Within this jurisprudence, Sisvel v Haier clarifies the conditions under which implementers may rely on the Huawei protections. Some commentators argue that Sisvel departs from Huawei by increasing the availability of injunctions, echoing the earlier Orange Book Standard doctrine. That characterisation is overstated. Rather than contradicting Huawei, the German Federal Court of Justice emphasised that protection for implementers is conditional on genuine willingness to conclude a FRAND licence. By requiring timely, serious and unconditional engagement in negotiations, Sisvel addresses the risk of strategic hold-out, where implementers delay or avoid taking a licence while continuing to use the patented technology. The judgment reinforces, rather than undermines, the effectiveness of injunctive relief as an enforcement tool against unwilling licensees, consistent with IPRED’s objectives.
This case law also points to why restricting injunctions would be economically counterproductive. Damages alone compensate only for infringement of individual patents and cannot address the realities of portfolio-wide licensing, where a single implementer may use hundreds of patented standards across a product line. Without the credible prospect of an injunction, infringers have little incentive to enter good-faith licensing negotiations at all: they can simply continue using the technology and treat any eventual damages award as a delayed, below-market licence fee. Limiting injunctions as a matter of EU law would therefore not rebalance enforcement; it would encourage efficient infringement (also called predatory infringement) and weaken the negotiating position of patent holders, including the SMEs, universities and research organisations whose primary commercialisation route is licensing rather than manufacturing.
PAE and SEP concerns Do Not Show an EU-Wide IPRED Problem
Some proponents of revising IPRED also point to concerns about patent assertion entities (PAEs), referring to entities that acquire and enforce patents primarily through licensing or litigation, non-practising entities (NPEs), referring more broadly to patent owners that do not themselves practise the patented technology, and SEP litigation. The Commission’s follow-up study reports differences in the way IPRED is implemented and applied across Member States and observes that national courts often apply only a limited proportionality assessment in patent injunction cases. That concern must be located correctly. The question is whether the evidence shows a systemic EU-wide problem caused by IPRED itself. On that question, the data points to a narrower conclusion: the concern is concentrated and forum-specific, and does not justify reforming the IPRED as a whole.
The UPC is the most important new forum-specific development and the available UPC evidence should be kept in proportion. Practitioner data from early 2025 already pointed to a meaningful NPE presence, with more than 70 cases representing roughly 14 per cent of the then-tracked UPC litigation dataset. That trend has continued: UPC infringement actions rose to 206 in 2025 (up 25.6 per cent on 2024), with NPE activity among infringement claimants increasing by 50 per cent and raising the NPE share of infringement claimants from 18.3 per cent in 2024 to 21.8 per cent in 2025. That updated figure is more specific than the earlier “14 per cent of all UPC litigation” estimate, because it concerns infringement claimants rather than all UPC proceedings. Broader UPC activity also continued to grow in 2026, with 170 new Court of First Instance cases recorded through April 2026. The better conclusion remains that UPC NPE activity creates a forum-specific risk requiring vigilance, especially in infringement actions, but does not by itself prove an EU-wide IPRED enforcement problem.
However, the increase in NPE activity remains too narrow and too indeterminate to justify a claim of systemic enforcement failure. That evidential caution is reinforced by the instability of the claimant categories themselves. An NPE is a broad descriptive category capturing any patent owner that does not practise the patented invention, while a PAE is a narrower and more policy-sensitive category, generally referring to entities that acquire patents from third parties and assert them for licensing revenue. These categories should not be treated as interchangeable. Nor should the fact that an entity does not practise the invention be taken, by itself, to undermine the legitimacy of its patent rights.
PAEs may acquire patents from original inventors or operating companies, as illustrated by Unwired Planet’s acquisition of patents from Ericsson, and the legal entitlement attaches to the patent itself rather than to the business model of the current owner. Accordingly, while the identity and conduct of a PAE may be relevant to a fact-specific proportionality assessment, a PAE’s right to protect its patents against unlicensed infringement is no less legitimate than that of any other patent owner. Any concern should therefore be directed at abusive enforcement behaviour rather than at weakening the underlying patent right. Given the different business models of universities, technology transfer offices, research organisations such as Fraunhofer, TNO, VTT and imec, spinouts, individual inventors and commercial assertion entities, broad claims of systemic enforcement failure cannot safely be inferred from aggregate NPE or PAE activity alone.
Moreover, the available data does not support that broad premise for a systemic EU-wide PAE or NPE problem. Although NPE activity before the UPC has increased, that development should be placed in context: NPE litigation before national courts appears to have declined, meaning that the overall increase in European NPE litigation remains modest. More importantly, PAE litigation remains predominantly a US phenomenon, while the European picture is more limited and concentrated. The 2024 NPE litigation data indicate that the ten most active NPEs account for 72 per cent of NPE-initiated cases in Europe, and that EU NPE infringement actions are concentrated against a small number of larger companies. This suggests that any concern is targeted and forum-specific, not diffuse or systemic. The appropriate response is therefore better monitoring of high-risk fora and abusive enforcement strategies, rather than a general legislative restriction of injunctions across the IPRED framework.
The scale of the underlying data reinforces this point. Looking at NPE-initiated patent infringement filings between 2018 and 2023, the US accounts for 10,632 cases, compared with 205 across four major European jurisdictions for which comparable data is available: Germany, France, Italy and the Netherlands combined – roughly 2 per cent of the US volume over the same period. That scale difference weakens the premise that Europe faces a PAE-driven or NPE-driven injunction problem comparable to the one often associated with US patent litigation.
The European data also show concentration rather than systemic spread. Of the 205 NPE infringement cases filed across Germany, France, Italy and the Netherlands between 2018 and 2023, 187 were filed in Germany alone, amounting to approximately 91 per cent of the four-jurisdiction total. Directive-level reform should respond to a directive-level problem; these figures instead suggest a geographically concentrated pattern in one long-established patent litigation venue, not a diffuse enforcement problem across the internal market. Germany’s own amendment to § 139, discussed earlier, which introduced a statutory proportionality exception while preserving the patent owner’s claim to compensation, further weakens the claim that NPE activity in Germany shows that IPRED itself requires broad legislative amendment.
Figure 1: NPE infringement cases filed in the US and four major EU jurisdictions, 2018–2023
Source: Author’s calculation based on Clarivate, 2024 Non-Practicing Entity Global Litigation Report, Figure 1C, using Darts-ip case-law collection data. EU figure includes Germany, France, Italy and the Netherlands.
The appropriate IPRED response should therefore be limited. PAEs, NPEs and SEP-related concerns may properly inform the case-by-case proportionality assessment, especially where the facts show a real risk of disproportionate injunction leverage. But they should not be converted into a general presumption against injunctive relief, nor justify weakening enforcement rights for all non-manufacturing patent owners, many of whom play a legitimate role in technology transfer and innovation. The evidence shows, at most, that assertion activity is concentrated in particular fora, that some venues may be more attractive to certain enforcement strategies than others, and that SEP/FRAND disputes raise distinctive remedial considerations. It does not show a systemic EU-wide IPRED problem requiring broad legislative restriction of injunctions.
Conclusions
The case for IPRED reform remains unconvincing. Although proportionality is an important principle, it is not a mandate for weaker IP enforcement. Rather, proportionality already operates within the existing legal framework as a fact-sensitive principle to be applied where it is properly raised and supported by the circumstances of the case. A better approach confirms that injunctive relief is the ordinary remedy following a finding of infringement, while leaving courts to address proportionality through existing procedural and substantive mechanisms where a serious proportionality objection is advanced. The UPC’s role is particularly important here. Because it offers centralised remedies across participating Member States, its decisions will shape the practical balance between effective enforcement and proportionality in Europe. It is precisely for that reason that its case law should be allowed to develop before the EU rewrites IPRED’s remedial framework.
The new IPRED debate should therefore, rather begin from a simple premise: proportionality is already part of EU IP enforcement law, and the case law – from Huawei through Sisvel – shows a framework that is already developing structured ways to assess the availability of injunctive relief in light of the facts of each case. The real question is not whether courts may consider proportionality, but whether the evidence justifies a legislative restriction on injunctive relief across the EU. It does not. The statistical evidence relied on for reform does not hold up to scrutiny. The real PAE concern is concentrated and forum-specific rather than systemic. The increasing emphasis on rebalancing enforcement in this debate reflects a policy preference for limiting injunctive relief in favour of alleged infringers, rather than a response to any demonstrated structural defect in IPRED.
Reopening the Directive on this basis would risk unsettling a remedial framework still being interpreted by national courts and the UPC, and would risk diluting IPRED’s original objective of high and effective protection twenty years after the Directive set out to achieve it. IPRED’s central compromise remains the right one: strong and equivalent protection of IP rights across the internal market, subject to safeguards against disproportionate and abusive enforcement.
One response to “Reassessing the Case for IPRED Injunction Reform: Proportionality, Evidence and European Patent Enforcement”