A discussion on recent developments in the field of SEP/FRAND

Dr. Schön, you are the presiding judge of the 7th Civil Chamber of the Munich I Regional Court. We follow your case law on the topics of FRAND and long-arm jurisdiction with interest. Recently, specifically in late April/early May, there were several decisions issued within a very short period of time by your court and courts in the People’s Republic of China and the United Kingdom in the global dispute between ZTE and Samsung.

Could you share your views on these various decisions?

I would like to try to contextualize our judgments in relation to the decisions of the other courts. In doing so, I can, of course, only refer to publicly available sources. It is also important to emphasize that in patent disputes involving international companies, courts in different countries have always dealt with specific partial aspects of a case.

Let me first define the context in which our decisions and the decisions of the other two courts should be viewed.

The Munich I Regional Court is currently probably the busiest patent infringement court in the Western world. In 2025, our two patent infringement divisions received over 330 new cases. This presents us with major challenges, because we have set ourselves the goal that over 95% of cases should conclude with a written first-instance decision within one year of the complaint being filed.

Currently, we schedule proceedings for newly filed cases 11 to 12 months after the complaint is received, and we do not distinguish between smaller patent infringement cases and major cases involving SEP/FRAND issues. We have optimized our internal preparation processes insofar as to enable us to maintain this timeframe on a permanent basis.

As presiding judge, it is important to me that plaintiffs’ counsel know that we are committed to adhering to this timeframe. This allows attorneys to advise their clients with certainty about how a proceeding will unfold. This predictability of procedure is an important aspect that distinguishes our court.

I would like to explain this to you in context.

As a starting point, one must consider the established case law in the Federal Republic of Germany, which has focused solely on the conduct of the parties, particularly the implementer. The requirements for a willing licensee were set very high, and as a result, the FRAND defense was never successful before German patent infringement courts.

When I took over as presiding judge of the 7th Civil Chamber in 2023, I found that the nature of the proceedings had changed. Cases involving an obvious hold-out have become rarer, and in many instances, the parties—sometimes with good reason—disputed only the amount of the license. Often, the problem stemmed from the fact that the patent holders had not submitted any license agreements. The implementers then argued that they could not reasonably be expected to enter into a license agreement without knowing what their competitors were paying. In some cases, the reasons why the patent holders did not provide license agreements were not readily understandable. However, one reason was also that it was not foreseeable how courts would handle license agreements.

This development meant that, in our view, it was no longer possible to achieve a fair outcome in SEP proceedings solely through negotiated settlements. As early as 2023, we therefore began making initial calculations in FRAND proceedings. However, these decisions went largely unnoticed because the reasoning in the judgments was redacted in large parts to protect the parties’ confidentiality interests. In the second half of 2024, we then resolved to draft our judgments in such a way that the underlying principles could be read in context.

After we made this decision, there was a very large series of settlements, which meant that from September 2024 and well into 2025, we had no suitable cases to publish our standard of review for the FRAND assessment in SEP proceedings. At the same time, our appellate court—the Munich Higher Regional Court (OLG München)—first issued a preliminary ruling and then a judgment on FRAND, which led to uncertainty among legal practitioners. We interpret the decision to mean that the Munich Higher Regional Court generalized an exceptional case and thereby set unrealistic standards.

We then issued a preliminary opinion in case 7 O 64/25—i.e., ZTE v. Samsung—in July 2025, in which we presented our principles in a comprehensive form for the first time. We then further elaborated on these principles in the decisions ASUS I (7 O 5007/25), ASUS II (7 O 4102/25), and RENAULT (7 O 7655/25). The most recent decision, ZTE/Samsung, concerns a special case in which the principles were nevertheless applied. Ultimately, however, the drafting of this decision demonstrated that the developed concept can lead to good results even in complex case scenarios.

We understand the Federal Court of Justice’s decision in FRAND III to mean that the standard of review has not changed compared to the FRAND I and FRAND II decisions. The Federal Court of Justice follows an approach that focuses on the parties’ conduct of negotiations, and so far, no decision has been made on exactly how a patent holder’s offer should be reviewed.

It is therefore the task of the lower courts to develop the legal framework for this issue and to issue initial rulings. These rulings will eventually, likely in 5 to 7 years, reach the Federal Court of Justice and become the subject of its decisions. The Munich Regional Court I has taken on a pioneering role in this regard due to the actual volume of cases. This is an interesting task because it involves an area of law in which courts from different countries influence one another.

The starting point is the question of whether we should address the patent holder’s offer in the context of a FRAND objection raised in patent infringement proceedings. And this depends on whether we view the implementer as fundamentally willing to license. If an implementer—within the framework of the negotiation process required by the ECJ decision in Huawei/ZTE—is clearly not interested in negotiating a license, then we see no reason to evaluate a plaintiff’s offer.

The situation is different if negotiations regarding the conclusion of a license agreement are actually taking place between the parties. In that case, it must be determined whether these are genuine negotiations or merely pretextual negotiations. To determine this, we have developed the concept of external and internal willingness to license.

By external willingness to license, we mean that the implementer is not engaging in an obvious hold-out. We assess this on the basis of a purely formal review. As a general rule, if the undisputed portion of the defendant’s offer has been paid and, if applicable, security in the amount of a one-year license has been provided, a review of the patent holder’s offer is required.

The review of the patent holder’s offer concerns the internal willingness to license. As part of this review, we determine whether the plaintiff’s license offer falls within the FRAND range or not. If the plaintiff’s offer falls within the range, then the implementer has no reason not to accept the offer. In this regard, it is important to understand that we do not determine an absolutely valid rate, but only ever establish a FRAND range.

As we understand it, the plaintiff’s offer is generally reviewed based on comparable license agreements. The result obtained in this way is then verified using a standardized top-down calculation to determine whether it is reasonable.

The problem lies in determining exactly what constitutes appropriate comparable license agreements. In this regard, we are of the opinion that the patent holder can determine which agreements to submit as comparable license agreements. If agreements with licensees are not submitted, this must generally be justified. The reason for this understanding is that we believe a patent holder should not be bound by unfavorable agreements once they have been concluded. On the other hand, however, it is also not FRAND if only a few agreements with small market participants are submitted in order to enforce a rate against a global corporation. Ultimately, it is the task of the judges to decide whether comparability actually exists.

In the ASUS II and RENAULT decisions, we examined the comparable licenses in great detail. In ASUS II, we found that the patent holder pursues and enforces a uniform licensing program. In RENAULT, we determined the license rate appropriate in this case by using several comparable license agreements. In this respect, slightly different assessment methods were applied. In the ZTE/Samsung decision, we concluded that there are no suitable comparable license agreements. Incidentally, this is the main difference to the decision in the United Kingdom. There, an agreement that was not introduced in our proceedings was considered a comparable license agreement.

In our view, the top-down approach should serve to verify whether the result obtained through comparable license agreements is appropriate overall. In this respect, it is a reasonableness test.

The top-down approach can take on particular significance in two scenarios. First, when the requested rate is so low that even a top-down analysis—assuming all calculation factors are set in favor of the implementer—leads to the conclusion that the patent holder’s offer is FRAND. We encountered this scenario in ASUS I. Second, when there are no comparable license agreements. In that case, the top-down approach serves to determine an appropriate license fee. We encountered this scenario in ZTE/SAMSUNG.

The real problem with the top-down approach is that one must determine the amount to be paid for the use of a specific standard in a device. There are various calculation methods in this regard, and our Chamber has opted for a standardized per-unit price. In the ZTE/SAMSUNG decision, we assumed that a price of $170 should be applied to a mobile phone, of which 8% is to be paid for the use of all mobile communication standards. As a result, this amounts to US$13.60 for the use of the device’s main functionality over the entire lifetime of the device. This is the AGGREGATE ROYALTY BURDEN (ARB). We believe this is a very low estimate that strongly accommodates the implementers’ interest in a low licensing rate.

Based on this calculation, the owner of a larger portfolio is therefore entitled to a royalty of $0.136 for every percentage point they hold in the standard.

We recognize that this approach can be criticized. There is room for debate regarding how the share of the total standard is determined or how valuable the portfolio is.

From our perspective, however, this can be left aside because the US$13.60 figure is very low and already accounts for these uncertainties. An ARB of US$25 to US$30 could likely also be well justified with appropriate factual arguments, at least when it comes to higher-end mobile phones.

The two decisions address slightly different directions. In ASUS I, a member of the patent pool filed a lawsuit, and we argued that the pool rate did not fully exploit the portfolio’s licensing potential. If such an offer is not accepted by the implementer, the implementer cannot be considered a willing licensee.

The ASUS II decision concerned Nokia’s offer, and in that case, the defendant argued that the offer was too high compared to the HEVC pool from Access Advance. The background is that the pool offer is very inexpensive. Less than $1 is required to cover approximately 90% of the standard. Thus, the licensing potential is not fully exploited.

We do not know how pool rates are determined. However, there is a tendency for them to be very low. In this respect, it is likely that the first preliminary injunctions in the SEP sector could be issued with regard to pool rates.

First, it is important to realize that we examine the FRAND defense within the context of our patent infringement proceedings. We assess whether the absence of a license agreement lies within the sphere of the plaintiffs or within the sphere of the defendants. If certain conditions are met, then we determine whether the plaintiffs’ offer at the time of the oral hearing falls within a FRAND range or not.

We take a market-based approach. It is up to the patent holder to determine what kind of license they wish to grant for their intellectual property rights. The boundary is the requirement to act in accordance with FRAND principles, and the patent holder has some leeway. FRAND addresses how this leeway is defined. On the one hand, an offer must be fair and reasonable. By this we mean that, in absolute terms, it must reflect what the patent holder has contributed to the standard. On the other hand, the offer must also be such that no market participants are discriminated against.

Furthermore, every offer has a time component. What constitutes a reasonable license fee cannot be answered the same way at every point in time. Those who invest in a company very early on may be able to negotiate very favorable terms for a long period of time. Furthermore, inflation adjustments must be taken into account. However, the behavior of an implementer during license negotiations is also significant. In other words: If a party resists obtaining a license beyond the normal extent, this results in additional costs for the patent holder, which must be compensated by a higher license fee.

We have already addressed this in several decisions. We assume that patent holders and implementers must be able to dispute the specific terms of a major license agreement. The starting point is the consideration that, in principle, there is a rough balance of power between patent holders and implementers during negotiations.

At first glance, one might see this differently, because patent holders can exert pressure on implementers through injunctive relief. However, we believe that such a view overlooks an important economic aspect, namely that implementers sell their products and thereby generate revenue. In other words: They have the money! And that money only shifts to the patent holders’ side if the implementer actively decides to pay. That is why implementers have the advantage, especially at the start of negotiations. A quick promise to pay can often secure favorable terms.

The cases that go to court are mostly those where the implementers did not negotiate skillfully and failed to take advantage of this leverage. When, shortly before an injunction is issued, implementers now say that the pressure is too great, they overlook the fact that this is, at least in part, a self-inflicted situation.

Against this background, we consider it appropriate to state that the option to file a lawsuit seeking an injunction is a legitimate cause of action in disputes over a license agreement.

Now to rate-setting proceedings! It is clear to everyone that no party can unilaterally determine which court is competent to set a global rate for all worldwide sales. Exceptions might apply, at most, where the respective defendant or the standard-setting organization is headquartered. If an implementer nevertheless decides to initiate a rate-setting proceeding, then an almost inevitable consequence is that this decision will affect the level of the rates.

I believe that “past release” is significant in two respects.

On the one hand, it concerns whether a patent holder can claim compensation for past use, and the other question is whether they must claim it. The second aspect is particularly important when considering comparable license agreements.

The starting point is that users of new technologies enter the market without having obtained comprehensive licensing. This is permissible because, especially in the early stages, energy and staff must be used for building sales structures and developing products. There may simply be no time left to deal with licensing.

On the other hand, there are the patent holders, who also have limited resources and must strive to secure product suppliers as licensees. This raises the question of whether they should set up their own licensing program or join a patent pool. The latter would have the advantage of being simpler, but the disadvantage of generating less licensing revenue.

This complex situation can result in a significant time lag between the start of production and the conclusion of a license agreement. There is fundamentally no apparent reason why the patent holder should forgo compensation for past sales. In this respect, there is generally no obligation to waive compensation for past sales or to grant discounts. However, the Chamber also believes that there are limits. The rationale is that the implementer is no longer able to pass on license fees to the purchasers of the products for past sales. In fact, this is therefore always a decision that depends on the case-by-case analysis.

The question of how to handle situations where parties have entered into a license agreement in which the past has not been settled must be assessed entirely differently.

Let me give you an example:

If the implementer sells the same number of products each year from 2015 to 2025, but the first license agreement was concluded from 2020 to 2025 and provides for a rate of 10 cents, then one could argue that the actual rate is 5 cents because the past has also been covered.

However, we consider such a view to be incorrect because it does not reflect reality. In such situations, the past is typically set aside to pave the way for the future. This is an economic decision that must be accepted by the courts. I believe that this issue in particular represents one of the greatest differences between the courts in Germany and the United Kingdom.

You are asking about the concept of determining the FRAND range as set out in the Renault decision, and I can understand that it sounds very broad at first if we assume that the licensing range between the cheapest conceivable license and the most expensive conceivable license can be three times as much. When considering these figures, it must be made clear that these are extreme positions. As a rule, the range will be much narrower. The key criteria are the size and scope of the license and the behavior of the implementer.

The cheapest license will likely be a contract renewal with a very large company covering multiple standards, where there were no issues whatsoever during negotiations. The most expensive license, on the other hand, is one concluded after years of disputes between parties in various jurisdictions, where the primary reason for the lack of agreement lies with the implementer. This is because, in such a case, the license must also compensate the patent holder for the additional effort.

I would like to illustrate this using the ZTE/SAMSUNG decision. In that case, we applied a 30% discount due to SAMSUNG’s high volume of units. However, we also applied a 10% surcharge because, in our view, Samsung is resisting the conclusion of a license to an excessive degree. In particular, we consider filing a request with the standardization body ETSI and initiating the rate-setting proceedings in the United Kingdom to be unacceptable means in the fight for a reasonable license rate.

Exactly. The focus is on determining the median value. We determine this median value by evaluating comparable license agreements, where available. In the RENAULT decision, we specifically examined the agreements with other automakers and then compared the model range and the size of the individual other automakers with Renault. Based on that, we then determined a median value.

It is important to note that in Germany we do not set rates, but rather examine whether the plaintiff’s offer falls within the FRAND range. We determined the mean value and then examined where the parties’ offers place. If the implementer’s offer is very far from the average value, then the patent holder may also maintain a slightly higher offer. For only then can the patent holder still grant discounts during the course of subsequent negotiations. As a consequence of this understanding, the same offer may lose its FRAND status if the implementer takes concrete steps toward the patent holder and the latter does not respond or does not respond appropriately.

However, I must also reiterate at this point that we, as the Chamber, do not have the impression that the patent holders’ offers are too high. Rather, it is usually the case that the implementers try by all means to drive down the amount of the license payments.

By “geographical adjustment,” we mean the application of different rates for the same product category in different countries or regions because the patent holder’s portfolio is validated to varying degrees.

Geographical adjustment is a fairly simple means of significantly reducing claims in a mathematically transparent manner. It was the defendants’ main argument in the RENAULT and ZTE/SAMSUNG decisions. The arguments differed in the two cases. In RENAULT, it was argued that the license rate for the defendant had to be lower because the company did not sell passenger cars in the U.S., and in ZTE/SAMSUNG, it was claimed that an adjustment was necessary because ZTE had not fully validated the portfolio in every country in the world.

I believe that to answer the question of whether and how a geographical adjustment should be made, one must be aware of what forms the basis for calculating a royalty. As far as we can tell, there are two approaches here. First, a party’s contribution to the development and further development of the standard, and second, compensation for the global portfolio of patents. This distinction is somewhat complicated, so I would like to explain it in more detail.

The situation is clear-cut when there are no intellectual property rights either at the location where the chipset is manufactured or at the point of sale. In such cases, no license fees can be charged. However, this generally applies only to holders of very small patent portfolios. In most cases, patent applications are filed across the globe.

Companies engaged in development will have the majority of their patents and patent applications in the countries where they are headquartered, because that is where the priority-establishing applications are filed. Subsequently, applications are filed in the most important markets. In less important markets, only the most important and strongest patents will be filed. Otherwise, it would be necessary to file and maintain all patents in every country in the world to obtain the same license everywhere.

In the ZTE/SAMSUNG case, we had the situation where ZTE held most patents in the People’s Republic of China (over 3,000) and only about 12% (over 400) of the patents were validated in the Republic of Korea. In Europe, a total of approximately 30% (over 1,000) of the patents were validated. Against this backdrop, the question arose as to how ZTE’s share of the mobile communications standard should be calculated. It was undisputed that the total number of patents in the standard amounts to approximately 60,000.

The Chamber is of the view that the share is calculated by referring to the total number of ZTE’s patents. The rationale is that the focus is on the contribution to the further development of the standard, rather than on a pool license for the individual patents and patent applications.

It is true that ZTE’s expert also made a geographical adjustment. However, that adjustment was largely based on actual sales.

The Chamber, on the other hand, bases its calculation on a standardized per-unit price of US$170. And we believe that by applying such a low value, geographical differences are sufficiently captured and taken into account.

It may be necessary in future proceedings for patent holders to also present their research and development expenses in order to determine the value of a specific portfolio.

Ultimately, we are entering new ground with our decisions. There is no established case law yet on exactly how to evaluate license agreements and license offers. We expect that the parties will present new arguments in future proceedings. Decisions by foreign courts are also an important source of information. Building on this, we will continue to develop our case law.

I believe that the jurisdiction of the German national patent infringement courts has expanded significantly because of the European Court of Justice’s BSH-Hausgeräte decision. In patent infringement proceedings involving the territory of other member states, we no longer need to stay the proceedings with respect to the invalidity defense. This also sends a signal to non-European countries.

Jurisdiction is derived from Art. 4(1) and Art. 8(1) of the Brussels Ia Regulation. Art. 7 of the Brussels Ia Regulation plays a subordinate role. The Lugano Convention applies accordingly.

An interesting question will be whether, under Article 8(1) of the Brussels Ia Regulation, parent companies headquartered in other European countries can be sued for a worldwide patent infringement. Our Chamber is of the view that this should be possible.

The more important question is how, in cross-border proceedings, the applicable national law of the respective state is incorporated into the specific proceedings. In proceedings concerning Member States of the European Union, our Chamber believes we can rely extensively on presumptions because the legal framework is highly harmonized by the European Patent Convention and the Enforcement Directive. For non-European states, this is far more difficult and can ultimately only be decided once suitable cases arise. With regard to the People’s Republic of China, it may be of interest

that Chinese law was heavily modeled on German patent law in the 1980s. This may help simplify access to that foreign law.

Module 1: German Courts' Role in FRAND Rate Setting and Global Jurisdiction

• Rate Setting: Do German national courts still tend to handle only SEP infringement actions and injunction matters, while being reluctant to adjudicate SEP royalty rate disputes? Will German national courts be prepared to determine SEP licensing rates on a case-by-case basis in the future? If so, what would be the geographic scope of coverage?

• Global Jurisdiction Precondition: Do German national courts hold that they have jurisdiction over global SEP licensing disputes only where the parties have explicitly agreed to request the court to determine global royalty rates?

• One-way vs. Cross Global Rates: In the ZTE v. Samsung Frankfurt antitrust case earlier this year, the court made a rough calculation of the global one-way rate. Does this mean German courts will in future cases be more inclined to decide on one-way or cross global rates/terms? Will such a tendency be linked to the assessment of whether the parties’ conduct complies with FRAND?

Module 2: Injunctions, the Proportionality Principle, and Assessment of Willingness

• Negotiating Fault Assessment: Do German national courts assess the parties’ negotiating fault solely by strictly following the Huawei/ZTE framework, or do they also take into account the parties’ respective offers and counteroffers when evaluating willingness?

• Proportionality and Injunctions: In light of the recent Samsung v. ZTE case (where an SEP holder failed to obtain an injunction based on willingness), will the German principle of proportionality be fully applied when assessing FRAND compliance? In German SEP infringement cases, does the so-called “infringement found = injunction” judicial principle truly exist, or is it a misunderstanding on the part of some commentators? If it exists, has this principle been comprehensively revised?

• Disproportionality Objection: Do you think the disproportionality objection against an injunction could be successful in certain SEP cases in the future?

Module 3: Interim Licenses, Security Deposits, and Jurisdictional Clashes

• Attitude Toward UK Interim Licenses: What are the views of German courts regarding the UK interim license regime, especially the recent trend where SEP holders proactively seek it? Will German courts consider introducing this approach in the future?

• Security Deposits vs. UK Approach: What are the general criteria for requiring an implementer to post a security deposit in Germany, and what is your opinion regarding the differences between the UK interim license and the German security deposit approach? Are security deposits and non-refundable payments necessary steps in the FRAND process, or do they have the effect of “curing prior unwillingness”?

• Timing and Pressure: If the licensee actively participated in the negotiations and acted correctly but failed to provide a security deposit in a timely manner solely due to financial pressure (e.g., failing to pay a security deposit when the counteroffer was rejected, but providing it immediately after litigation commences or shortly before the hearing), how should willingness be assessed?

• Jurisdictional Tension: What is your view on the current tension between UK courts and the UPC/German courts regarding global rate setting and interim licenses? What would be the possible solutions to mitigate such tension between the different jurisdictions?

Module 4: Cross-Licensing and its Special Security Requirements

• Complexity of Cross-Licenses: Why are cross-license agreements particularly difficult to assess in FRAND cases? What legal or economic factors distinguish them from ordinary bilateral licenses, and can they truly serve as reliable comparables?

• Mutual Security in Mutual Suits: In a situation where two SEP holders sue each other and the outcome is likely to be a cross-license agreement with one party acting as the net payer, how does this affect the requirement to provide FRAND security and make a partial payment? Should the net payee also provide security and make an undisputed payment to the net payer, and how should the amount be determined?

Module 5: FRAND Methodology, Past Releases, and Expert Evidence

• Evaluating FRAND & Market Practice: How should courts evaluate whether a license offer is truly FRAND? What role should comparable licenses, market practice, and economic expert evidence play?

• Unpacking Past Royalty Payments: Is a discount for past royalty payments an obligation or a right of the patent holder? Is there any difference between unpacking a “comparable license” and calculating the royalty component for a new license? How should past royalty payments be unpacked—do discounts affect the rate, or do they affect the past time period for which royalties have been paid?

• Rate Calculation Tools: Are geographic adjustment factors and an implied ARB useful tools for assessing offers or comparables? What is your take on the Broadcom v. Renault case, in which the Munich Court proposed that the highest permissible value of a FRAND range is three times the lowest value, or a 50% deviation from the mean? Does this apply only to this case, or is it intended as a general rule?

• Other Factors & ECJ Guidance: Do standstills and covenants have an impact on the calculation of the FRAND rate? Is there a need for more structured guidance from the European Court of Justice on how FRAND rates should be assessed?

Module 6: Trial Procedure, Confidentiality, and Emerging Sectors

• Trial Models, in reviewing the preliminary injunction application: What is your opinion on the trend among some Brazilian courts (specifically those in Rio de Janeiro) over the past two years of requiring a positive, expert-backed verification of FRAND compliance before granting a preliminary injunction? This practice was further refined in the 2025 Sun Patent Trust v. Great Wall Motor case, where the judge bifurcated the preliminary assessment by appointing two separate neutral experts: a technical expert to evaluate patent essentiality and infringement, and an economic expert to independently verify the FRAND compliance issue.

• Trial Procedures and Confidentiality: Could you outline the general SEP trial procedure in German courts, particularly how they facilitate the exchange of confidential information (such as comparable licenses)?

• Transparency Limits: How much transparency can courts realistically require in FRAND disputes, and what are the challenges regarding confidential negotiations? Does limited transparency make it harder to ensure consistent outcomes? When German courts refer to third-party licensing agreements in SEP infringement actions, do they apply the same transparency standards as those adopted by UK courts in the rate-setting process?

• Emerging Sectors & ADR: How do German courts address rate calculation challenges in emerging sectors (e.g., streaming media) that lack comparable samples and top-down core data like ARB? What are your views on arbitration awards versus court litigation for SEP disputes?

Module 7: Licensing Levels, UPC Long-Arm Jurisdiction, and Parallel Litigation