Standard Essential Patents (SEPs)
How Should In-House Counsel Advise Their Clients on SEPs?
As a PhD student at Cornell working on wireless communications many years ago, I associated standard essential patents (SEPs) with the cellular industry. Later I realized that the universe of SEPs is much larger.
Consider video technology. Whether it's video streaming, video conferencing, or digital television, video is compressed into smaller sizes for efficient transmission over the network. This is done using video codecs. Short for "compression-decompression," codecs compress the video while maintaining acceptable image quality. When a compressed video file is played back, the codec decodes it, restoring it to its original form for viewing. Popular video codecs include H.264, H.265 (HEVC), VP9, and AV1. These multimedia standards are patent protected - some of which are essential to the standards.
So when companies such as Netflix, YouTube, or even Twitter develop video streaming products, one business consideration in selecting which video codec to use is the cost of patent licenses associated with that codec. As in-house counsel, how do you advise the business on the cost of patent licenses to the codec? 1
In technology areas with established patent pools, e.g., Via LA Licensing for MPEG video codecs or Avanci for wireless technologies, the answer is less complicated. These pools attempt to provide "one-stop licensing" for SEPs by aggregating patents from multiple patent holders and offering licensing solutions to manufacturers, developers, and service providers who wish to implement the standards in their products or services. For example, Via LA offers standardized licensing programs that cover all essential patents within a particular MPEG standard. This simplifies the licensing process for licensees, as they can obtain licenses for multiple patents through a single agreement with Via LA.
The answer is more complex in the absence of patent pools (and sometimes even with them because a patent pool may not comprehensively cover all relevant SEPs). Fundamentally, SEP licensing boils down to an SEP holder asking an implementer to license its patents at a certain royalty rate. From an implementer’s perspective, a big challenge is lack of transparency between the parties on essentiality (what patents are truly essential) and the rate others are paying (what is RAND?). When the parties cannot agree on a rate, potential litigation awaits. These days such litigation is frequently global in nature. The parties fight in Europe, India, China, and increasingly Brazil, in an attempt to find a favorable forum.
In a recent panel on SEPs at the Silicon Valley IP Forum, Stanford Law School, Jorge Contreras, Bolko Ehlgen, Dan Lang, Michael Lee, and The Honorable James L. Robart provided apt commentary on the state of SEP litigation across key jurisdictions.
US. The US is not a favorable jurisdiction for SEP litigation because under eBay it is difficult to obtain injunctions. In the present SEP litigation climate of jurisdictional competition where different forums are vying to set global FRAND rates, and anti- anti- . . . suit injunctions have become a thing, Judge Robarts notably issued the world’s first anti-suit injunction when he enjoined Motorola from enforcing in the US the injunction it obtained against Microsoft in Germany. Microsoft Corp. v. Motorola Inc.
UK. Around 2020, the UK emerged as an appealing jurisdiction for SEP holders due to its willingness to set a global FRAND royalty rate, as seen in the Unwired Planet case. However, recent cases in 2023 tempered this attractiveness. In Interdigital v. Lenovo and Apple v. Optis, while the UK court determined a global FRAND rate, it was notably lower, aligning more closely with the implementers' demands than with the SEP holders' expectations.
Germany. This forum is popular for SEP litigation, primarily due to its distinct feature of granting automatic injunctions upon finding patent infringement, unlike the United States where injunctions are not guaranteed.
UPC. This is a new forum that is largely untested with respect to SEP litigation. There are currently 120 cases pending at the UPC; 17 are SEP related. Some of the earliest SEP decisions came from the German division of the UPC (staffed by German judges). As would be natural, these early findings were consistent with the German SEP approach. For example, the Munich division issued the first big preliminary injunction at the UPC in 10x Genomics v. NanoString. The injunction was reversed on appeal, so the UPC Court of Appeal may emerge as a counterpoint to the German approach over time.
Further changes are afoot in Europe with proposed SEP regulation. The regulation aims to facilitate SEP licensing by reducing information asymmetries between SEP holders and SEP implementers. It intends to do this through:
SEP registration. SEP holders will be required to register their SEPs with the EUIPO (not the EPO because the EPO is not an EU institution). SEPs that are not registered will not be enforceable. The point below clarifies whether the patents are actually standard essential.
Essentiality checks. Over declaration of essentiality is common. In negotiating a FRAND rate with an SEP holder, even well-resourced implementers find it challenging to determine which patents are actually essential. They typically hire specialists who charge as much as $10,000 per patent. Under the proposed EU SEP regulation, the EUIPO will check whether registered patents are essential. EUIPO will hire specialists for this task but it is unclear how many checks will be conducted per year. The essentiality checks will be non-binding.
FRAND determination. The EUIPO will develop a process to determine FRAND royalty rates. The process will be mandatory before an SEP holder or an implementer can go to court, although the outcome will not be binding. It's anticipated that the procedure will mirror the German SEP approach, despite historical reluctance from German courts to set FRAND rates. While the EUIPO decision won't be binding, the FRAND rate that is determined will establish a framework for negotiations between parties. Over time, this process could enhance predictability in FRAND rate negotiations for implementers.
It is worth observing that the above procedures will apply to future standards. Current standards such as those relating to 5G or WiFi 6 will be unaffected.
From the perspective of an in-house attorney advising the business, another interesting question is: How should SEPs be valued for new technologies? The panel broke this down into three categories.
1. New application of existing technology
Example: wireless technology traditionally used for cellular applications being incorporated into cars or IoT devices.
Given that cars are an established platform, the simple solution is to get an Avanci license. Avanci offers a single license that covers multiple patents related to cellular connectivity standards like 2G, 3G, 4G, and 5G. Effectively, a very large percentage of the stack is licensed in one transaction (e.g., Avanci has 57 patent owners in its Avanci 4G Vehicle marketplace including Ericsson, Qualcomm, InterDigital, and Nokia). The process is fairly streamlined. In addition, given that automakers like Audi, BMW, Ford, General Motors, Honda, Mercedes-Benz, Nissan, Porsche, Renault, Stellantis, Toyota, Volkswagen and Volvo have all obtained Avanci licenses, you are more certain of getting the same royalty rate as your competitors. However, Avanci rates were established through the threat of injunction rather than true negotiation.
2. The technology is an incremental improvement over existing technology
Example: 4G to 5G and subsequent improvements.
Should SEPs for the new generation of technology command higher FRAND rates? There are arguments on both sides. On one hand, patents directed to the improvements are arguably complex, which could justify a higher price tag. Conversely, since the new generation is an incremental improvement rather than a 0-1 step change, some argue that they should cost less.
3. Completely new technology
Should the associated SEP IPR policy be royalty-free or RAND? In the scenario where only a small number of companies (3-4) are in the new technology space and their objective is to make their products interoperable, a royalty-free policy is sensible. Many web standards, such as those promulgated by W3C, are royalty-free (even though there are many web companies now).
While cellular SEPs command RAND rates, the panel observed that non-zero royalty rates don’t seem essential for innovation. Web standards are largely free and they have continued to develop over time.
Copyright © 2024 Shrut Kirti. Published here under a Creative Commons Attribution-NonCommercial 4.0 International License.
The views expressed herein are the author’s own and do not reflect any positions or perspectives of current or former employers.
VP9 and AV1 are royalty-free standards.