Proportionality and FRAND in France - Paris appeal court upholds anti-anti-suit injunction but first instance court refuses PI in IPCom v. Lenovo

IPCom is one of Europe's most active – and infamous – patent assertion companies and is currently involved in litigation against various parties in Germany, the UK, and France. The litigation resulted in a string of interesting decisions: in the UK, the High Court held that one of IPCom's patents was valid and infringed by Vodafone, but held that it might invoke the Crown use defence [here, Katpost forthcoming]; in France, IPCom was awarded an anti-anti suit injunction last November [Katpost here]. And in Germany, IPCom is allegedly requesting a staggering EUR 12 billion in damages for past infringement by Nokia [here].

Now, two further decisions have been handed down by the French courts in the proceedings between IPCom and Lenovo. The anti-anti-suit injunction granted last November was upheld on appeal [here, in French], while in separate proceedings a preliminary injunction (PI) was denied because the court considered it disproportionate [here, also in French]. Both decisions are quite thought-provoking and deserving of a closer look [with thanks to PatentMyFrench for first posting about the decisions and uploading them!].

The anti-anti-suit injunction

As some readers may know, the anti-anti-suit injunction is a legal innovation that made headlines last summer, when the Munich Regional Court granted such an injunction to Nokia in proceedings against Daimler [Katpost here, upheld on appeal, see here]. In the UK, the High Court reportedly also granted an anti-anti-suit injunction, also in proceedings brought by IPCom [see here; the decision does not seem publicly available]. Around the same time, IPCom was also granted an anti-anti suit injunction in France [here], which has now been upheld by the Cour d'Appel.

The anti-anti-suit injunction is a response to the practice of some courts, notably in the U.S., to grant anti-suit injunctions which bar a party from bringing proceedings in other jurisdictions. According to the Court of Justice of the European Union (CJEU), anti-suit injunctions are incompatible with the Brussels I Regulation [see C-159/02 Turner v. Grovit] but in practice a party must obey the injunction or risk being held in contempt by the U.S. court. Hence, parties facing a potential anti-suit injunction in the U.S. have resorted to asking European courts to issue an anti-anti-suit injunction to force the opposing party to withdraw the claim for an anti-suit injunction in the U.S.

So too here. When negotiations with IPCom stalled, Lenovo initiated proceedings before the U.S. District Court for the Northern District of California, requesting it to set the terms for a global FRAND-licence. When IPCom hit back with an infringement action, Lenovo asked the District Court to order IPCom to withdraw these proceedings, and order it to refrain from filing other infringement actions elsewhere in Europe. IPCom responded by requesting anti-anti-suit injunctions in London and Paris, both of which were granted.

At first instance, the Paris court held that: (i) there was no lis pendens obstacle to hearing IPCom's request as proceedings to set a FRAND-rate were held not to have the same subject matter as proceedings for an anti-anti-suit injunction; and (ii) anti-suit injunctions issues by non-EU member states can only be recognized in Europe to enforce contractual jurisdiction or arbitration clauses (i.e. to keep a counterparty to the forum they agreed on).

On appeal, the reasoning was different. The appeal court held that it had jurisdiction to grant the anti-anti-suit injunction to avert harm that would be suffered by IPCom in France if it would be barred from enforcing its patents there [par. 45]. It added that the French preliminary relief judge would also have jurisdiction to grant the injunction because it is ancillary to the jurisdiction that the French court has on the merits to hear an infringement claim [par. 47].

Interestingly, the court upheld the anti-anti-suit injunction not because it was incompatible with the European legal order, but rather because it would violate IPCom's rights under the French Intellectual Property Code, the European Convention of Human Rights, and the Charter of Fundamental Rights [par. 57]. Put briefly, these legal instruments afford to IPCom the right to act against infringements of its patents and to protection of its (intellectual) property, and guarantees it a fair trial.

The appeal court considered it relevant that the duration of the anti-suit injunction could potentially be very long in light of the possible length of U.S. proceedings, whereas IPCom's patent was set to expire on 15 February 2020. The anti-suit injunction would thus have effectively amounted to a deprivation of IPCom's property [par. 58].

To this Kat, the reasoning of the appeal court holds more appeal (no pun intended). By justifying the grant of the anti-anti-suit injunction in terms of IPCom's fundamental rights, the substantive justification of the grant becomes more persuasive and the content of those rights is further developed.

On the other hand, and as I wrote before, I think that there are certainly situations where an anti-suit injunction might be appropriate, pending litigation to set a global FRAND-rate. If an injunction is granted in parallel infringement proceedings, that might well induce the implementer to accept non-FRAND settlement terms and thereby moot the FRAND-proceedings: indeed, that is exactly what the first instance court suspected IPCom was trying to achieve, as we will see shortly. For this reason, Carr J granted an anti-suit injunction in the FRAND-proceedings between Conversant and Huawei [here]. As the case law on anti-anti-suit injunctions develops, it will be interesting when (if ever) courts will refuse it in favour of the pending FRAND-proceedings.

Good things come in pairs

IPCom's request for a preliminary injunction

As the appeal concerning the anti-anti-suit injunction was pending, IPCom moved quickly by requesting a preliminary injunction against Lenovo's French subsidiaries and its distributors. The PI, however, was denied because the court considered it disproportionate in light of all the circumstances of the case, citing article 3 and recital 22 of Directive 2004/48 (the Enforcement Directive). This decision is extremely interesting because it analyses in detail the proportionality of a preliminary patent injunction, requested by a non-practicing entity. To my knowledge, it is among the first decisions of this kind in France.

Lenovo cum suis argued that an injunction would force them to remove nearly all Motorola phones from the French market, that this would cause them very substantive financial and reputational damage, and that this was not justified in light of the short remaining lifespan of the patent (which, again, expired on 15 February). They pointed out that IPCom does not itself practice the patent and could therefore be adequately compensated with damages if infringement were established. But their damage would be hard to quantify and compensate, especially because the average lifespan of a phone is two years so that Lenovo and Motorola would lose potential consumers for a time long after the expiry of the patent.

For its part, IPCom argued that, instead of engaging in licensing negotiations, Lenovo employed stalling tactics by bringing a case in the U.S. IPCom argued that denying it an injunction would make it lose credibility not just vis-à-vis Lenovo, but also in relation to other potential licensees. This would seriously devaluate its patent portfolio, which is its primary business asset.

The court sided with Lenovo. Emphasizing specifically that the Enforcement Directive requires that remedies be proportionate, it held that the proportionality of requested measures must be evaluated on a case-by-case basis, in particular by examining the harm caused to parties by granting or denying it. It accepted Lenovo's reasoning and considered it relevant that IPCom does not itself practice the patent and so could be compensated with damages. The damage to Lenovo, on the other hand, would be "exorbitant" since it would be felt long after expiry of the patent. Under these conditions, the grant of the patent would give IPCom an "undue advantage".

Reasonable minds can disagree on how the proportionality analysis turned out in this case, but it is significant that the court so explicitly engaged with the Enforcement Directive's proportionality requirement. This signals a willingness to take a more cautious, case-by-case approach to injunctions, which is an important step forward. True, the case concerned a PI where courts traditionally enjoy discretion to balance parties' interests. But the court's reasoning is not based on this discretion, but instead on the proportionality principle, which applies with equal force in merits proceedings.

Several questions remain. At some point the court suggested that it is up to the patentee to show not just a high likelihood of infringement, but also that allowing the infringement to proceed pending merits proceedings would cause it serious harm not capable of being redressed with damages. That would be a significant shift in the burden of proof, reminiscent of the approach taken in the U.S. following eBay v. Mercexchange, but unknown in Europe.

One may also wonder why the court did not simply apply the framework set out by the CJEU in C-170/13 Huawei v. ZTE. After all, it is not clear that a proportionality test must be applied on top of the requirements of this decision when requesting an injunction on the basis of a FRAND-patent. Maybe the court felt that this assessment could not take place in PI proceedings – as the District Court of the Hague recently held, see here – but it did not say anything to that effect. Either way, proportionality in Europe patent law is a development still in its infancy, so no doubt these questions will be addressed when the time comes.