Monday, April 3, 2023

In draft EU SEP regulation, appeals are not even an afterthought: what if owners or implementers of standard-essential patents disagree with decisions?

This is a follow-up to the following three posts:

In the entire draft regulation, one very important word appears only once, and just in a footnote: "appeal". That footnote references a recent judgment by the Court of Appeal of The Hague. There are no search results for "appellate".

One would normally assume that appellate jurisdiction would be addressed at least in the recitals. Either you allow appeals or you don't, but even if you don't, you should make it clear instead of leaving it to the courts to determine whether they have jurisdiction. Only the draft impact assessment mentions appeals, but just in the form of input the EC received on two questions, only one of which is directly relevant to the draft regulation. One of those recommendations is legally--and laughably--untenable at best and academically dishonest at worst, and the other even contradicts the EC's approach (as I'll show further below).

The question of access to justice is key for all intents and purposes, including that the ultimate "appeal" in a figurative sense would be an annulment action before the European Court of Justice by one or more EU Member States. Having read the draft regulation, I could put together a relatively long list of issues that might be raised in that scenario, which is anything but unlikely (it may almost be a given should anything similar to the current draft be passed into law). Rule-of-law issues such as access to appeals courts could be raised at that stage.

The authors of the draft regulation don't even take appeals of decisions by patent offices and courts into account. For example, Article 26 says that a declared-essential patent will be removed from the SEP Register in certain events including but not limited to:

  • "invalidation of the patent by a competent authority" (i.e., a patent office or court),

  • "judgment that the patent is not essential to the standard for which a registration is made"


  • "negative result from the essentiality check pursuant to Article 34(3) and Article 35(7)."

Invalidations or non-essentiality judgments can be appealed, which potentially takes years. The draft regulation does not explicitly say that only a final decision (not appealed or after exhaustion of all appeals) will result in removal. It also doesn't clarify that any information stored in the SEP Register about such decisions would have to clearly indicate whether a decision is still under appeal, so as to avoid confusion or even misinformation.

Even an essentiality check performed under the envisioned regulation would not necessarily be final. Under some procedures that the Commission would determine afterwards, there would be a "peer review" of a subset of essentiality checks. Only in this context does the draft regulation explicitly rule out further "review". Recital 28 says:

"Unless a SEP is the subject of a peer review, there will be no further review of the essentiality check results."

Apart from that being only a recital, there could be an argument over whether "review" also means "appeal." The narrower meaning would suggest that there won't be another review by one of the EUIPO-appointed evaluators. But what if there is a legal basis for an appeal?

It is a major problem that the identity of an evaluator will not even be disclosed after the evaluation (Art. 30(8)). I can see why one would not want the evaluators to be influenced while they're doing their work. But if a decision raises issues, a conflict of interest or lack of expertise in the relevant area may have been a factor.

For an institution that likes to lecture the whole world--at times even the United States--about democracy and the rule of law, it's remarkable what the Commission's draft SEP regulation envisages.

Interestingly--and inconsistently--the draft impact assessment addresses CJEU review in connection with the "merits of potential candidates for the Competence [C]entre". It notes that EUIPO decisions are subject to review by the CJEU (which is contrasted with hypothetical alternatives such as the EPO, which has its own Boards of Appeal) and SDOs. So the decisions made by those EUIPO employees should be subject to review, but the potentially even more important ones by independent "evaluators" and "conciliators" should be off limits for any appeals court?

Between the lines, the argument is made that those essentiality checks are non-binding and can be challenged in court (i.e., in the infringement cases). That doesn't convince me at all. If those essentiality determinations had zero relevance, there would be no point in making them in the first place. The obvious purpose is to give them some persuasive weight. There is definitely prejudice to a SEP owner from a false negative and to implementers from a false positive. Take preliminary injunction proceedings, for instance: under the time constraints, courts may be persuaded to order a PI over a false positive or to deny one over a false negative. There is also an asymmetry: a single erroneous determination by a single EUIPO-appointed evaluator could require a party (be it a SEP owner or an implementer) to convince multiple courts (for instance, the UPC plus courts in Spain, Poland, and other non-UPC countries) that the evaluator made a mistake.

Those essentiality checks of individual SEPs are used to determine an essentiality ratio for each SEP holder. That, in turn, informs any FRAND determination--I'll get to the appealability of those in a moment.

There are overwhelmingly strong arguments for the right to appeal such determinations--and there can definitely be no argument for disallowing appeals under circumstances that involve breaches of procedural rules. Let's face it: if someone got bribed, would there be a way to disallow all legal recourse? Obviously not.

A document that is titled "Impact Assessment Report" and meant to accompany the legislative proposal contains the following footnote 283:

"Allowing for appeals to the assessments of randomly sampled patents produces few benefits, and significant costs – appeals are likely to correct a significant share of random assessment error, which is a relatively benign and inconsequential error for purposes of assessing firms’ relative portfolio size (false positive and false negative random errors tend to cancel each other out). Appeals are however likely to exacerbate, rather than correct over-confirmation bias, as negative assessments are significantly more likely to be appealed. By reducing random error and increasing bias, appeals make the assessments of firms’ relative portfolio sizes less reliable. Source: Baron, J., Arque-Castells, P., Leonard, A., et al., Empirical Assessment of Potential Challenges in SEP Licensing, European Commission, DG GROW, 2023."

This is preposterous. It's an insult to human intelligence. It tells us that the document has the word "empirical" in its title but it may just be a non-scientific work-for-hire. It will be interesting to see the full document on May 4, but that passage alone makes me doubt that I won't disagree with significant parts of it.

The claim that "false positive and false negative random errors tend to cancel each other out" is misplaced for two reasons:

  • There is no scientific evidence for this, and I doubt we'll find any in the May 4 study. They certainly cannot have any empirical evidence for the future process that would be administered by the EUIPO. The way the process works, and the selection criteria for the future evaluators, could certainly lead to a bias. I also doubt that there has been any scientific study on whether courts of law did or did not agree with third-party essentiality checks that delivered positive or negative results.

  • The authors would at least have to address a question that the Commission's summary of their findings does not talk about: the statistical risk of a single SEP holder being seriously prejudiced because of misassessments. What totally discredits the paper (unless the Commission misrepresented something, which I doubt) is the fact that those essentiality checks will impact even the smallest SEP holders. Theoretically, a company with a single SEP may find that one listed in the SEP Register, and it might have to be reviewed. Even if it's a small portfolio of, say, 5 SEP families, it is statistically highly unlikely that a false negative affecting one of them will be canceled out by precisely one false positive.

Do those economists discuss the statistical likelihood of deviations for samples of different sizes in their study? We'll see soon.

Does the Commission want to rely on economists who don't even appear (again, I need to see the final document) to apply statistical logic? What about the rule of law? That is not for economists to opine on even if they do apply common sense and statistical logic.

What's also ridiculous is the assertion that "negative assessments are significantly more likely to be appealed." That is, in fact, the way the whole patent prosecution process works: while third parties may submit prior art and oppose grants, most of the time a patent examiner is faced with a patent attorney who will keep fighting until the patent is granted or no more reasonable arguments can be raised. But that doesn't mean that one can just disallow appeals.

There could also be decisions on costs that private parties would want to challenge.

Let's also talk about appeals of FRAND determinations. There will be aggregate royalty determinations for entire standards and FRAND determinations involving the relationship between one SEP holder and one implementer at a time.

What I said about essentiality checks and ratios applies here as well: while those determinations are non-binding, there could be breaches or even wrongdoing involved, and a party (be it a SEP owner or an implementer) might have to challenge such determinations in parallel or sequential proceedings in multiple courts. Also, those determinations can pave the way for preliminary injunctions (if good for the SEP owner) or for their denial (if an implementer benefits from an error).

Those inter partes FRAND determination proceedings are the closest thing to arbitration that one can find in the draft regulation. The whole process reads like "arbitration mutatis mutandis". But any professionally-crafted arbitration agreement addresses appealability. Sometimes the scope of appealability is wider, sometimes it's narrower, but it raises rule-of-law concerns if appeals are impossible even in the event of a major breach.

On page 230, the draft impact assessment interestingly acknowledges the following:

"Arbitration is also seen as unfit to contributing to a wider understanding of FRAND, including because the proceedings are confidential. Arbitration may not compel discovery of third parties. Arbitrators are faced with the absence of clear rules, which will result in unsatisfactory decisions that are not subject to appeal."

The difference between the FRAND determination proceedings envisaged by DG GROW and arbitration is that the latter results in an award that is enforceable. I'm not denying that the two types of proceedings can be distinguished. I'm not saying that those FRAND determination proceedings would be arbitration by any other name. But the fact that the courts of law are free to disagree doesn't mean that a party--be it a SEP owner or an implementer--will not be prejudiced, even if only indirectly so, by erroneous results.

In Art. 45, the draft regulation lays out a process for objections to a FRAND determination. But that is just the equivalent of a motion for reconsideration by the very same panel. That falls short of what the rule of law requires.

The draft regulation wants the conciliators to apply the ECJ's Huawei v. ZTE case law as they make their FRAND determinations. But what if an important question concerning the interpretation of Huawei v. ZTE comes up in the process?

The largest ones of those FRAND determinations would be made in connection with multi-billion-euro disputes, and in those large cases, the costs (especially if one includes the parties' costs for legal, economic, and technical advisers) may be in the millions--possibly tens of millions--of euros. Now imagine that a party that disagrees with the result on substantive and/or procedural grounds appeals such a decision to the EU General Court. With all that's at stake, and with all of the money that the parties will have invested by then, would the EUGC say it has no jurisdiction because the EUIPO just issued a non-binding opinion?

I'm far from done raising and discussing the issues that I've already identified. There'll be more posts in the coming days and weeks. The draft regulation and the draft impact assessment need a major overhaul. Based on that footnote 283 of the draft impact assessment, it's doubtful whether that "empirical" study can serve to justify any of what DG GROW intends to do here. It might be better to cancel that May 4 event rather than present a "study" that will invite harsh criticism. For now I'm still prepared to give the authors of that study the benefit of the doubt, but only because Commission staff may have mischaracterized their position on the appealability of essentiality checks...