Updated IP Report on draft SEP Regulation

A little over a month after the draft of an SEP Regulation of the European Commission was leaked, the Commission released an amended official draft on April 27, 2023, comprising many clarifications and amendments. To provide readers with some background, we refer to our earlier assessment of the initial draft. A comparison shows countless amendments ranging from small corrections to the addition or deletion of entire paragraphs. Nonetheless, the basic concept of the draft has not changed, and the impact of the amendments might be surprisingly small. Hereinafter, we will provide a more detailed analysis with a focus on three main pillars: SEP register, essentiality checks, and FRAND determination.

1. Scope of the Draft SEP Regulation

Several important clarifications regarding the scope of the regulation were made: In accordance with the Regulation, the determination of aggregate royalties and FRAND terms shall refer to global rates and conditions, and the Commission may exempt certain standards from the Regulation in its entirety1

The first clarification is remarkable because outside the UK and China no courts have claimed the authority to determine worldwide FRAND terms. It raises the question why the EUIPO is believed to be competent to decide about FRAND royalties for patents and products/services even outside the EU.

Likewise, the impact of the second clarification could potentially be very important. The Commission mentions that, for certain standards “such as standards for wireless communications”, licensing practices are well-established so that there is no need for regulatory intervention2. If “wireless communication” standards are exempted from the SEP Regulation, its scope is materially reduced and key sectors of SEP licensing remain unaffected. The criteria for determining a well-established licensing practice and thus exempting a standard from the Draft SEP Regulation are not entirely clear. However, if “established licensing practices” are the decisive benchmark, then this would probably also apply to standards for which a large portion of SEPs is licensed by patent pools with hundreds of licensees and dozens of licensors.

Another important clarification is made in Article 1 which defines the subject matter and scope of the Regulation. It is now stated that the Regulation only applies to SEPs for which the holder has issued a FRAND commitment3. This means that any SEPs held by entities that did not participate in the standard-setting or not declared are not covered by the Regulation. Since the non-registration of an SEP results in a loss of material rights by its holder, such an exemption is reasonable and necessary because a company that did not participate in the standard-setting often does not know or only finds out later that certain patents are standard-essential. A significant related issue is the extent to which FRAND commitments made by prior holders are attached to the SEPs and thus are included in potential assignments. In any event, this impairs the intended transparency of the envisaged SEP register.

What has not changed is the definition of what an SEP is under the Regulation. Notably, even patents covering optional parts of standards are encompassed as “essential to a standard4. This broad definition raises questions with respect to the underlying competence of the Commission for a regulatory intervention and puts the draft at odds with national case law. A patent that is implemented in an optional part of a standard cannot convey market dominance to its holder (except for exceptional de facto essentiality cases). Implementers may simply choose to not use the option. If a patent can be circumvented technically and commercially, related licensing terms cannot distort the market – which is cited as the underlying reason for the Draft SEP Regulation by the Commission5. Furthermore, at least under German law, without market dominance, the patent holder is not subject to any FRAND obligations.

2. Registration and Essentiality Checks

The amendments made to the provisions on the registration of SEPs and essentiality checks do not change the gist and principles of the leaked draft. In particular, the new Articles 29(6) and 32(2) are noteworthy.

According to Article 29(6), not only patent holders but also implementers may propose up to 100 SEPs annually for essentiality checks. This has the potential for exponentially increasing the workload of the EUIPO because every (online) shop selling electronic devices might be regarded as an implementer. So, many companies might propose 100 patents for essentiality checks each year, making such checks an impossible task. To avoid this, one could argue that an implementer, defined as “person that implements […] a standard“6 is only the entity making the unit that implements all features of the patent. However, that would exclude many downstream companies such as smartphone makers, etc. since SEPs are largely implemented on chip level and the case law of many jurisdictions accepts that implementation occurs at various levels of the commercial chain.

Article 32(2) is, from a legal perspective, a clear improvement because it makes a peer evaluation of an essentiality check mandatory if the SEP holder disagrees with the result. Previously, such review was still left to the discretion of the competence center; making it mandatory brings it closer to an actual legal review of the decision. But concerns remain with respect the fundamental rights to fair trail and the right to be heard. The evaluator and the peer conducting the review are anonymous and there is no judicial review of these administrative decisions, even though they can have a huge impact.

Likewise, another flaw we pointed out in our initial assessment of the leaked draft was not remedied. The essentiality checks are still limited to one patent per family. While this certainly reduces the workload for the competence center, it also very much limits the transparency provided by the register and allows circumvention attempts by patent holders. A particular patent in a family may or may not be standard-essential, whereas other family members may be the opposite. Therefore, the register might include many patents that are in fact not essential. SEP holders may try to hide real SEPs behind non-SEP family members to circumvent the requirements of the Regulation.

Also, the issue removal of patents from the register after they have been found to be non-essential by the EUIPO is not satisfactorily resolved. Here, the Regulation seems to allow the patent holder to still assert the patent in court as essential. Nothing in the draft indicates that the courts are bound by the essentiality checks of the EUIPO. Even if they were binding, this would aggravate the concerns that there is no judicial review of the decisions made. So, ultimately, patent holders might not mind a negative essentiality check, as it spares them FRAND determination proceedings at the EUIPO prior to litigation, and might have an incentive to play with respective arguments.

Finally, a truly major point has not been resolved. The Regulation is silent on the standards for claim construction. In the EU, the law and case law on claim construction is not yet harmonized which means there is no common legal basis for determining essentiality. 

3. Aggregate Royalty and FRAND Determination

The provisions on the determination of aggregate royalties and FRAND terms are the most important and most controversial parts of the Regulation. Here, some significant amendments were indeed made:

SEP holders can notify the EUIPO about what they consider an adequate aggregate royalty for a standard. This request has to be made “jointly”, i.e., by at least two patent holders. The information shall be published by the EUIPO but beyond this, the consequences remain unclear. The draft SEP Regulation does not state whether and to which extent the request and/or the EUIPO’s evaluation is binding. It can reasonably at most bind the SEP holders having made the request, but initial views could change over time and in view of arguments made. From a practical perspective, it is unclear why any SEP holder should make such a request. And even if it is made, it does not provide any certainty or transparency for implementers since it is not binding for other SEP holders and can be revised any time.

A request from implementers for a non-binding expert opinion on aggregate royalties would probably be more relevant. Then again, since it is not binding, this could remain a complex yet academic exercise.

The FRAND determination process stipulated in Art. 34 et seqq. is mandatory prior to certain types of litigation and will thus be the much more relevant issue in practice. However, it is important to note that the process is only mandatory prior to specific proceedings, specifically proceedings in which

  • the SEP holder asserts claims for infringement and/or
  • he implementer asserts a claim for determination or assessment of FRAND terms and conditions

prior to the FRAND determination by the EUIPO7. This limited scope results in a procedural imbalance at court level: implementers can file declaratory proceedings for non-infringement and SEP holders can file actions seeking to determine FRAND terms and conditions without a prior FRAND determination by the EUIPO and without the other party having the procedural right to file respective counterclaims or defenses.

Further concerns relating to predictability and legal certainty remain unresolved in view of the procedural framework for the actual FRAND determination by the EUIPO, although an important amendment was made in this regard: From the outset the envisaged FRAND determination resembles court proceedings. It starts with a written phase during which the parties submit arguments, documents, expert opinions, witness statements, etc. Then, upon request of a party or if the conciliator considers it necessary, an oral hearing is held. Finally, after a nine-months window at the latest, the conciliator determines the FRAND terms. Knowing the importance of procedural rules and their application for a FRAND determination in court, it is astonishing to see that the regulation remains completely silent on how the FRAND determination procedure shall actually be conducted. For instance, there are no rules on the burden of proof. There are no provisions either on which forms of evidence are admissible and what the respective evidentiary value is. The conciliator seems to have very wide discretion without any tangible limits. A few examples:

  • the conciliator may hear witnesses and experts, provided “that there is time to consider such evidence”8,
  • the conciliator may request the production of documents or other evidence9 or
  • examine publicly available information as well as confidential information of other FRAND determinations and non-confidential information provided to the competence center10.

Against this background parties are well-advised to not commit to the results of the FRAND determination up-front although they could in view of a significant amendment made to the draft:

Once a request for FRAND determination is filed with EUIPO, the responding party shall, within 15 days, indicate whether it agrees to the FRAND determination and whether it commits to comply with its outcome11. If the responding party agrees to comply, the requesting party will be asked whether it also commits to comply to the results of the process and, if it does, the result is binding for both parties12. However, if either the responding or the requesting party refuses to commit to the results of the FRAND determination, the other party can choose whether it wants to continue or terminate the procedure13. Once the FRAND determination procedure is terminated, the road to the courts is open14. This mechanism which, at first sight, seems balanced due to its reciprocity could play out very differently for SEP holders and implementers:

If an implementer requests a FRAND determination, the responding SEP holder will most likely not commit to comply with the result – given the uncertainties of the procedure. If the implementer in turn chooses to terminate the FRAND determination procedure, the SEP holder can start infringement litigation in court. If, however, the implementer chooses to continue the FRAND determination procedure, the SEP holder is barred from bringing litigation until the procedure is finally terminated. In this context the SEP holder might even prefer to not engage in the FRAND determination to be able to argue in court that a one-sided result came out of it which has no precedential value whatsoever for the litigation.

If, however, the SEP holder requests the FRAND determination procedure, the situation is different: The implementer may well refuse to commit to its result (for the lack of predictability explained above) which means that the SEP holder could then simply terminate the FRAND determination procedure and move on to court litigation.

In addition to that, Art. 47 hints at another problem: If (prior to or during a FRAND determination procedure before the EUIPO) court proceedings are brought based on an SEP that “in substance has the same claims” as the SEP subject to the FRAND determination in a country outside the EU15, the FRAND determination procedure can be terminated upon request by the other party16. Thus, if a party, prior or in parallel to the EUIPO FRAND determination procedure, brings respective litigation anywhere outside the EU, both FRAND determination procedures can continue in parallel. But in contrast to the procedure under the SEP Regulation, the decisions in FRAND determination proceedings in China or the UK, for example, are binding. Thus, the party responding to a FRAND determination under the SEP Regulation can effectively torpedo this procedure by filing an identical request outside the EU, which will legally override the results of the EUIPO.

Finally, the question remains unresolved whether or not an SEP holder has to conduct a FRAND determination procedure for each and every implementer. This is important as it could either allow for discriminatory findings or for findings in which an implementer had no say. 

4. Conclusion

We appreciate that various clarifications and amendments were made to the initial draft. However, many issues remain unsolved and cause legal concern. We thus expect heavy lobbying to take place which might lead to either many further amendments or to a complete withdrawal of the Commission’s plan. The latter would not be a bad thing as SEP licensing be it out of court (which is the vast majority of cases anyway) be it in court in outlier cases – works well in practice without any need for regulatory intervention.


1 Recitals (5) and (8).
2 Recital (4).
3 Article 1(2).
4 Article 2 (1) and (2).
5 See Recitals (3) and (4).
6 Article 2(7).
7 Article 34(1).
8 Article 49.
9 Article 48(1).
10Article 48(2).
11 Article 38(2).
12 Article 38(3).
13 See Article 38(3) and (4).
14Article 34(5).
15 Article 47(1).
16 Article 47(2).



Tilman Müller
Attorney-at-Law (Rechtsanwalt), Certified IP lawyer, UPC Representative, Partner

Tilman Müller

Tilman Müller-Stoy
Attorney-at-Law (Rechtsanwalt), Certified IP lawyer, Commercial Mediator (MuCDR), UPC Representative, Partner

Tilman Müller-Stoy