The EU has embarked on a mission. A mission to reduce transaction costs in Standard Essential Patents ("SEP") licensing and litigation. With its latest proposal for SEP Regulation (COM(2023) 232 final of 27 April 2023) ("Proposal"), the EU Commission aims at the dissemination of technology for the mutual benefit of SEP holders and standard implementers. An ambitious claim. Can the SEP Proposal deliver? Not necessarily, I believe. Read more to find out why.
SEPs confer market power to their holders. SEPs cannot be avoided when selling standard-compliant devices such as, for example, 5G-enabled smartphones. To avoid the hold-up of standard implementers (i.e., market foreclosure), SEP holders, therefore, are under an obligation to inform the alleged infringer of the infringement complained about and to offer to license the SEP (or portfolio of SEPs) concerned on fair, reasonable and non-discriminatory ("FRAND") terms prior to seeking injunctive relief in court. Conversely, to avoid the hold-out of SEP holders (i.e., fair remuneration of patented technology), the alleged infringer must be willing to take a FRAND licence on whatever terms are in fact FRAND (Court of Justice of the European Union, decision of 16 July 2015, C 170/30, ECLI:EU:C:2015:477 – Huawei v ZTE).
The details are left, or have been left, I should say, to negotiations between SEP owners and standard implementers, subject to judicial review. SEP owners will need to show that their offer is FRAND and if they fail to do so, they will lose on FRAND i.e., no injunction will be issued. How they do it is up to them. Tested means include disclosure of comparable licence agreements with similarly situated companies, top-down analyses, third-party essentiality studies and total aggregate royalties, to name only a few. Implementers, on the other hand, will need to engage in negotiating the FRAND terms of the licence to be concluded and provide a FRAND counteroffer, as the case may be, in due time and, if they apply delaying tactics, they will get injuncted. No FRAND objection will be available to them.
A balancing of powers (and risks) that, with a limited number of exceptions, has proven to work well, following judicial guidance including from, most notably, the CJEU in Huawei v ZTE , the UK High Court in Unwired Planet v Huawei (High Court of England and Wales, [J. Birss], decision of 5 April 2017,  EWHC 711 [Pat]) and the German Supreme Court in Sisvel v Haier (German Federal Supreme Court, decision of 5 May 2020, KZR 36/17 – Sisvel v Haier I; decision of 24 November 2020, KZR 35/17).
The EU SEP Proposal
The EU Commission's SEP Proposal opts for a more regulated approach to SEP licensing and litigation. SEP owners will need to register their SEPs with a newly formed administrative body (a "Competence Centre" to be established within the EUIPO, the Union's Intellectual Property Office) and they will be barred from seeking court remedies for unlawful patent use, including injunctive relief, prior to having completed a mandatory FRAND determination process under the patronage of the EUIPO. Read in conjunction with a number of further provisions (and in the absence of practical evidence to the contrary to date), it appears the Proposal tips the balance towards standard implementers and raises transaction costs mainly for SEP owners rather than reducing them. In detail:
To support transparency in the SEP field, SEP owners will be required to register their SEPs with a Competence Centre to be established within the EUIPO within six months following the opening of the registration or the grant of the relevant SEPs, whichever is first. Failure to comply will mean that SEP holders cannot collect royalties and claim damages for uses and infringements that occur until the SEPs concerned are registered (see Articles 20 and 24).
SEP Competence Centre and Database
The tasks of the Competence Centre include, amongst others, to perform essentiality checks of registered patents and to build a database of opinions, reports and available case law including from outside the EU (see Article 3). The database shall also contain information on relevant standards, products, processes, services, and systems which implement the standard, SEPs in force in the EU, and SEP owners' standard licensing terms and conditions (i.e., standard licensing programs, if existing) (see Article 4).
Most of this information is already available to date. To ensure that standard-setting conforms with Art. 101 TFEU and avoid "patent ambush" (EU Commission, Horizontal Guidelines, C(2023) 3445 final; R. Schellingerhout/P. Cavicchi), patent owners are under an obligation to disclose patents that they believe are essential to the future standard ("declared-essential patents") to the relevant standard development organization ("SDO") prior to the standard being adopted. The information is available on the SDO's website. Numerous third-party consultancies offer essentiality checks, the results of which are usually made available to standard implementers by SEP owners at their own initiative to justify their rates (including in court as the case may be; see above).
Moreover, confirming that product a implements standard b does not seem to add much value – standard-compliant devices are typically marketed as such. Take 5G for example. The challenge is to judge whether patent claim x reads on standard feature y – and that's a responsibility the infringement courts are tasked with, not a database.
Finally, where standard rates exist, typically they are
published e.g., on the SEP owner's website. However, from
experience the cases that go to court concern complex, often
multi-element, business cooperation, and/or lump sum agreements
that do fit in standard rates (and that the standard implementer
agrees do not fit, seeking a better or, more neutral, different
Supporting transparency in SEP licensing is no doubt a noble goal but, in this case, it comes at high costs in terms of both time and money. Costs that are borne by SEP holders and implementers alike. SEP holders will have to pay a registration fee for their SEPs and implementers an access fee for the database. Implementers may argue that they will get independent third-party data in return. However, considering that such data, to the extent inconsistent with the SEPs own data, will be challenged in court, it appears the added value is marginal.
The cost of time is even higher. The Proposal includes a FRAND determination process of nine months (standard but can be extended) as a precondition to bringing a patent infringement case to court. The FRAND determination process is mandatory, and it can be initiated and completed even against the SEP owner's will and without it committing to the outcome by unilateral request from the implementer who, by doing so, will be shielded against litigation for the time the determination process continues (Article 34; Preamble at 34).
Nine months is a time in which in some jurisdictions, most notably Germany, a first instance decision would have been available to the parties already. There is no better facilitator for the conclusion of a FRAND licence than a court decision. If implementers are given the opportunity to unilaterally initiate a FRAND determination process, this will likely only delay the finding of an amicable out-of-court settlement as implementers will have no incentive to engage in negotiating the mutually acceptable FRAND terms of the licence to be concluded.
On substance, the FRAND determination process is focused on determining a total aggregate royalty burden ("ARB") for all SEPs covering a standard. However, in most cases I have been involved in, ARB was no decisive factor. When a new standard or standard generation is agreed, SEPs owners may express different views on reasonable ARBs. However, the differences are typically in the low single digit area and, from experience, less relevant to finding a court decision on FRAND. Moreover, where competing SEP owners "agree" on future ARBs (as the Proposal seems to suggest, see Article 15, Preamble at 16), these will likely be higher than under normal market conditions. In effect, this means that higher rates can be supported by SEP owners as being within a FRAND range of terms and conditions when in licensing negotiations or litigation with implementers.
Also, it should be noted that, if at all, ARB is only one of many elements used to determine a FRAND rate. Others include namely comparable licences and how these translate into the rate requested from the infringer at hand. ARB, by contrast, is a concept that assumes each SEP to be of equal value, which is why it is commonly referred to also as "licence pie theory". While this may be true if measured exclusively against each SEP's potential of blocking market access for standard-compatible devices, undisputedly some SEPs contribute more to the technical features, and hence commercial success, of a standard than others. It is for this and many other reasons that implementers may be willing to pay a premium for some SEPs but not for others – and that the holders of the first may rightly request to receive a "larger piece of the pie" without acting in breach of their FRAND obligations. There is no better indicator of FRAND terms and conditions than the prices agreed in previous market transactions.
Finally, the ARB concept assumes that the prices of standard-compatible end user devices are a given. In contrast, you could argue that if the final end price of a device does not allow for adequate (and additional as the case may be) remuneration of the technologies implemented in such a device, then maybe the price point set by the implementer is too low, rather than the royalties requested by the SEP owner too high.
On the procedural side, the FRAND determination process will be led by a "Conciliator" to be appointed by the Competence Centre (Article 39). Again, the Proposal fails to demonstrate how exactly this will serve its aim of "facilitating the negotiations and out-of-court dispute resolution that can benefit both parties" (Preamble at 31). Experience shows that SEP owners and implementers already make use of alternative dispute resolution mechanisms such as mediation and arbitration. The difference is that, under the existing law and practice, these mechanisms are run by private bodies (e.g., ICC), do not provide an incentive to delay the conclusion of a FRAND licence, and allow for in-court litigation to start or continue in parallel.
Determining the FRAND terms and conditions of the licence to be concluded is at the heart of every SEP dispute. Until today, doing so was left to negotiations between SEP owners and implementers, subject to judicial control. This approach is approved, including by the CJEU. It is in line with the principle of subsidiarity and, with only limited exceptions, has proven to work well.
The Proposal chooses a different, more centralised, and more regulated approach. It remains to be seen how the courts of the Member States will deal with the outcome of any EUIPO-led FRAND determination and whether they will at all feel bound to it. Different from e.g., the UK and the US, to date, most of the courts of the Member States have refrained from determining a FRAND rate. A decision on FRAND was found on other grounds including most notably unwillingness to license on the side of the implementer. Will SEP owners who are faced with an implementer unwilling to engage in licence negotiation prior to the conclusion of the FRAND determination process at EU level and unwilling to license on FRAND terms other than those determined by an EUIPO-appointed Conciliator be at all able to get a court injunction in the future? We don't know yet. What we do know, however, is that, if adopted in its current form, the Proposal will significantly change the way in which SEPs are being licensed and litigated in the EU in the future. Whether for the good or bad remains to be seen.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.