OPINION
A headline like the one above would go too far if I merely disagreed with something on which reasonable people can disagree. Unfortunately, what the European Commission’s (EC) Directorate-General for Competition (DG COMP) put out last month (September 11, 2025 ip fray article) as draft guidelines on patent pools and licensing negotiation groups (LNGs) contains a mix of
- potentially intentional misconceptions,
- unworkable ideas (such as an expiration date on group boycott), and
- an utter illogicality that unfortunately constitutes the legally most important part of the EC’s attempted rationalization of purchasing cartels called licensing negotiation groups (LNGs).
Two assumptions would, of course, be naive. First, if the EC really understood technology markets, Europe wouldn’t find itself in a position now where it does not have a single one of the world’s 25 most valuable companies. Second, when an antitrust watchdog is simply a department of a larger government body and the actual decisions are made by a vote among 27 politicians who know nothing about competition law, it is inevitable that the ones in the trenches will be given directions for political reasons, forcing them to scramble to rationalize the irrational and defend the indefensible.
For example, two years ago it was a key milestone for Microsoft’s $79B acquisition of Activision Blizzard that the EC approved the deal, but DG COMP imposed “remedies” that were clearly unsupported by its own findings. Its UK counterpart took more extreme positions, but at least they were internally consistent. Regrettably, DG COMP gets away with it too often. For example, the European Court of Justice (ECJ) upheld the “state aid” ruling against Apple in the Irish tax case, though the actual issue was one of the EU’s fundamental design flaws and not the equivalent of market-distorting subsidies. They sometimes bend the law and make up their own facts. It’s not principled. It’s politics. The reward the ones in the trenches get for sometimes doing what they actually know is wrong is a level of compensation that is low compared to Big Law, but very high, especially after taxes and considering pensions, for European public servants. Also, a number of them got recruited by Big Law or Big Tech, and not necessarily always (though probably sometimes) for their actual knowledge.
So let’s look at some nonsense they wrote in the LNG context (probably not because they really believed it but at the automotive industry’s behest) that has not been highlighted before, at least not on this website. The proposed guidelines are available here. Before we get to the crazy stuff, it’s important to clarify the context.
Para. 266 starts with the following sentence:
There is no inherent link between technology pools and standards, but pools often support, in whole or in part, a de facto or de jure industry standard (122).
Footnote 122 is even stronger (emphasis added):
See concerning the treatment of standards and the treatment of standardisation agreements in Guidelines on Horizontal Agreements, Chapter 7, cited in footnote 16. Technology pools typically operate licensing programmes that are structured around specific industry standards.
That means we’re now primarily talking about standard-essential patent (SEP) pools, and the first LNG to have received a comfort letter from DG COMP is all about SEPs, too (July 9, 2025 ip fray article).
In the part about LNGs, there are fundamental issues with the way DG COMP looks at market power. Market power starts with market definition, and that’s where they ignore distinct characteristics of the way in which patented technologies are implemented, actually used and commercialized (see Sisvel founder Roberto Dini’s September 29, 2025 Politico op-ed).
SEP licensors understandably dislike the reference to a “small but non-transitory decrease of the price offered for their technologies” (as a result of allowing LNGs) in para. 315. At first sight, it suggests that the EC’s agenda is to devalue SEPs. But in private conversation I’ve actually defended that wording because the context is market definition, and unambiguously so. A small but non-transitory price change is a hypothetical used to determine market power, which in turn helps define relevant antitrust markets.
Similarly, though for practical reasons not to the same extent, I’ve defended the safe harbor criterion set out in para. 326(g):
the licensing fees payable under technology transfer agreements negotiated through the LNG do not exceed 10 % of the sale price of the products incorporating the licensed technology
That is not governmental price-setting. The rationale for this must be that if aggregate licensing costs are below 10% of the product price, and if one assumes that an LNG gives its members a cost advantage of, for example, 5%, the net effect on product price is maybe half a percent or less, which does not distort competition in the product market. That still doesn’t make it a good idea. For example, how will this really work in practice if during negotiations it turns out that aggregate licensing costs surpass the threshold? But I can see the point in being less concerned about LNGs where there is no serious risk of a purchasing cartel putting non-members at a comparative disadvantage.
Para. 319, which is closely related to the one with the small but non-transitory decrease (para. 315), is the first of (only) two paragraphs discussing the potential effects on technology (i.e., patent licensing) markets. The other one, para. 320, talks about an unrealistic scenario of licensees imposing restrictions on license deals with non-members of their LNG that is never going to matter in practice.
After outlining the adverse effects on innovation that, for example, a group boycott can have, para. 319 says “such negative effects are less likely where technology holders have countervailing bargaining power” and gives the following examples:
That may be the case, for example, where the relevant technologies are essential to a standard that is widely adopted in relevant downstream product markets or for which there are no substitutes; where there are very few licensors and they hold large portfolios of technology rights, or where the LNG negotiates with a technology pool.
Let’s look at the parts left and right of the semicolon separately:
- “where the relevant technologies are essential to a standard that is widely adopted in relevant downstream product markets”: That is a question of market power that the ECJ addressed in Huawei v. ZTE. Given that the ECJ believed that it was solving the problem at the time, the EC would have to explain why the market power concerns the ECJ has already (from its point of view) resolved still exist and require a permissive approach to purchasing cartels. If that was the case, the EC would have to address it differently. It would not be a horizontal cooperation issue.
- “where there are very few licensors and they hold large portfolios of technology rights, or where the LNG negotiates with a technology pool” (emphases added): This. Is. Nuts. Here’s why:
As the ECJ held in Huawei v. ZTE, a single truly standard-essential patent (which obviously means it must be valid, too) confers market power. Nowhere did the ECJ indicate that there was any concern over a plurality of truly standard-essential patents giving its holder greater market power. ZTE, the implementer in that case, apparently did not even argue that the size of Huawei’s portfolio (already formidable back then) was of concern.
Motorola Mobility once (in)famously argued in an expert report filed with a court in a SEP enforcement case that “it takes only one bullet to kill.”
There is no way that a SEP pool has more market power because of the number of patents it contains. It either has at least one truly standard-essential patent, in which case the owner of that patent (rather than the pool) has market power, or it does not, in which case the number of patents doesn’t matter either.
That correlation between pool size and market power vis-à-vis an implementer is a fallacy, and probably just an intellectually dishonest attempt to defend the indefensible because some politician(s) told them to do the automotive industry a favor.
It’s a false equivalency:
- In a market where concentration on the supply side can result in artificial scarcity, that logic would work. For example, if automakers faced a single tire maker (or tire cartel) with an incredibly high market share, it might be difficult to source enough tires from other potential suppliers. SEPs, however, are not substitutive. A patent is a right to enforce, not a license to do something. Even if (which patent law doesn’t allow anyway) two patents covered the same invention, a license to one would not br a defense to the infringement of another.
- The availability of bilateral licenses alone negates any market power theory, be it over a single SEP or any number of SEPs.
- By contrast, if a SEP holder or a SEP pool faces an LNG with high market share in a distinguishable market segment, it is restricted in its ability to have separate negotiations with the various potential buyers, and the power of the purchasing cartel does grow with market share.
If there was any reason to be concerned about a pool’s market share, it would not be because of its share of SEPs on the given standard making it more expensive for implementers to secure a license. If anything, one would have to look at whether the market for patent pool administration services is competitive. That has nothing to do with leverage in negotiations between a pool and an LNG. In that regard, the ones affected by it would be licensors who believe they are not getting a fair share of the transactional efficiency gains realized, and no such complaints have ever become known. Also, even a monopoly is not unlawful if it resulted from merits-based competition as opposed to illegal monopolization or monopoly maintenance. If every administrator had the opportunity to form a pool for a certain standard, and if the winner takes it all, then the outcome is not attributable to restrictions on competition.
I’ve seen (literally) hundreds of SEP cases and related decisions. Never once has there been a serious issue because of a patentee’s portfolio size. Any market power discussion focused on the scenario of (at least) one patent being truly standard-essential.
Actually, if one thinks it through, the existence of a pool with a very high share of SEPs reading on a given standard makes it even less justifiable to allow a puchasing cartel. The hypothetical small but non-transitory decrease of the royalties offered by implementers when they jointly negotiate would be anticompetitive if it exceeded a fair split of the gains of trade, i.e., of the transactional efficiency gains achieved by a collective-licensing mechanism. But if most or almost all relevant SEPs are available from a one-stop shop, there already are transactional efficiencies in place that it is hard to achieve any significant further efficiency gain by allowing LNGs.
Theoretically, pool size would matter if the patents in the pool covered all viable alternatives (for example, all relevant standards of a different kind, and even that would only confer market power if neither non-standardized options, such as creating a proprietary charging system, nor bilateral licenses were available; or non-SEPs that cover alternative techniques and jointly make it impossible to compete in a market without infringing at least some non-SEPs in the pool and, again, if no bilateral licenses are available). But that is not what the EC had in mind. It says patent pools are typically about one standard at a time.