Sunday, October 27, 2019

Intel antitrust lawsuit takes aim at Softbank-owned Fortress Investment's patent aggregation, obfuscation, and litigation tactics

As this week draws to a close, I realize that the most important IP topic of the week was the role of financial engineering behind some of the most aggressive patent transfers and assertions. Huawei's counsel in the Unwired Planet v. Huawei case told the Supreme Court of the UK that Unwired's aggressive attempts to monetize former Ericsson patents amounted to "leveraged financial engineering." Earlier in the week, Intel filed an antitrust and unfair competition complaint in the Northern District of California against Fortress Investment and three of its subsidiaries (two of which are patent assertion entities named VLSI Technology and DSS Technology Management). Fortress Investment was in free fall until it was acquired by Japan's Softbank for $3.3 billion in 2017.

In its lawsuit filed on Monday (October 21, 2019), Intel alleges antitrust violations

  • either in the market for patents for high-tech consumer and enterprise electronic devices and components or software therein and processed used to manufacture them

  • or, in the narrower alternative, in the market for licenses to Fortress's aggregate portfolio.

Antitrust analysis generally starts with market definition. There must be a market in which someone has a dominant position and acts abusively.

Intel isn't saying that investments in patent licensing firms or large-scale acquisitions of patents by patent assertion entities would always be illegal, or that trollish litigation tactics raise competition concerns. The first couple of pages of the complaint discuss the policy implications of patent assertion entities (PAEs) at a generic level, but the specific case is about certain structural and behavioral characteristics of Fortress's industrial-scale patent acquisition and assertion business model.

Citing to what network technology company Sonus Network alleged in a case against one of Fortress's numerous shell companies, Inventergy, the complaint quotes that particular entity's CEO as telling Sonus that "Fortress[,] does not settle" in litigation and, in the absence of a license deal palatable to Fortress, Sonus would face "an IP bloodbath." The flowery language of that threat does not per se constitute an antitrust violation, and the complaint doesn't say or suggest so. It merely serves to illustrate how little Fortress's business model has to do with innovation and to what extent the business model is simply to create, and capitalize on, a patent reign of terror. Case in point, a particularly well-known Fortress entity, Uniloc, previously caught my attention because it shows up in the RPX Daily Litigation Alert very often as they've brought dozens of lawsuits against Apple and Google, as well as other defendants.

Fortress apparently sets up and shuts down patent assertion entities at an unusually high frequency. They transfer patents between them, sometimes as a result of subsidiaries being unable to meet their payment obligations to the holding company. Similarly, they just dismiss complaints in one venue to refile somewhere else. And when patent claims are found invalid, they sometimes come up with many dozens of amended claims that allegedly don't have any more merit (as they just add some meaningless terms to the claim language), but enable them to keep suing forever.

Again, vexatious and oppressive litigation tactics don't in and of themselves constitute anticompetitive conduct. The point I found particularly interesting from an antitrust angle is that Intel explains how Fortress systematically acquires, through different subsidiaries, patents covering alternative techniques so as to make it practically impossible to work around all of those patents without Fortress being able to allege (whether with or, more likely, without merit) some infringement(s) at any rate. I couldn't find the term "patent thicket" in the complaint, but that's the one that came to my mind when I read the related passage:

"23. Further, aggregating a massive portfolio of electronics patents allows Fortress and its PAEs to amass a range of patents that are both substitutes for and complements to one another. When a firm wants to build an electronic device, such as a smartphone, there are many ways to do so. Each alternative requires multiple technologies. However, the alternatives do not require the same combination of technologies. For example, Alternative 1 might require technologies A, B and C, while Alternative 2 might require technologies D, E and F. The technologies used for Alternative 1 (A, B and C) are complements: they are each needed to create the device using Alternative 1. Similarly, the technologies used for Alternative 2 (D, E, and F) are economic complements. The technologies comprising Alternative 1 are also a substitute for the technologies comprising Alternative 2, because the bundle of technologies used in Alternative 1 can be used as a substitute for the bundle of technologies used in Alternative 2.

"24. There are many possible permutations of complement and substitute technologies for electronics patents. For instance, Alternative 3 might require technologies A, C and D. In that scenario, the technologies bundled in Alternative 3 are a substitute for the technologies bundled in Alternatives 1 and 2 respectively; A, C, and D are complements in the production of Alternative 3; and technology D is a substitute for technology B. Technologies can thus be both substitutes and complements. If Alternative 4 used technologies A, B, and D, then B and D are complements for Alternative 4, and substituting D for B changes Alternative 1 to Alternative 3."

Another allegation is that Fortress requires companies to license numerous patents deemed meritless (so weak that they "never would have been asserted by their former owners") in order to license those that are not that weak. Package deals are common in many industries, and the allegation here is very much about Fortress's patent aggregation strategies. It's not about aggregation of the efficient kind where licensees would be presented with a one-stop solution: while the Fortress web of companies as a whole engages in large-scale patent aggregation, companies face royalty demands from numerous Fortress companies and are never offered a deal covering the patents held by all Fortress entities.

According to Intel's complaint, "Fortress and its PAEs foreclose the possibility—which existed before aggregation—that litigation can be an economic alternative to licensing patents." In other words, Fortress allegedly bases its monetization strategy in no small part on the nuisance value of meritless patent lawsuits that result in what I would call hard (i.e., legal fees) and soft (i.e., distraction of employees) costs to those forced to defend against Fortress's infringement actions.

The complaint mentions the following Fortress PAEs--note that any of those PAEs may itself have spawned numerous companies (in the U.S. as well as abroad):

  • VLSI Technologies allegedly discussed three alternative ways of helping NXP maximize its income from a part of its patent portfolio: Financing, Privateering, and Corporate Carve Out (an acquisition of a copany division along with its patents). Guess what--the chosen route was Privateering. Two years ago, VLSI asserted eight former NXP patents "against virtually every one of Intel's microprocessors ever sold since 2011" and sought $7.1 billion. That first case got stayed when PTAB IPRs were instituted against six of the patents-in-suit. Thereafter, VLSI brought a couple of Delaware cases, at least one of which also involved a multi-billion-dollar damages claim. But with injunctions not being realistically available in the U.S. (except from the ITC in the form of import bans), VLSI is also suing Intel in China.

  • DSS sued Intel as one of various defendants (electronics companies as well as retailers like Wal-Mart). Intel settled earlier this year, but presumably on very favorable terms as the patent claims-in-suit had been declared invalid by the PTAB.

  • Uniloc's dozens of lawsuits were mentioned above. To be precise, various Uniloc entities have so far sued Apple 25 times in the U.S. (Eastern and Western Districts of Texas), apparently mostly or exclusively over cellular standard-essential patents acquired from Philips, and over the course of only three months brought a total of 35 lawsuits against Google. That's 60 just between Apple and Google--and there have been more than 70 other Unioc infringement suits already.

  • Inventergy acquired many hundreds of patents from companies like Nokia, Panasonic, and Huawei, then sued Apple, HTC, and ZTE in the District of New Jersey and is seeking an ITC exclusion order (import ban).

  • IXI sued Samsung, BlackBerry, and Apple.

  • Seven Networks sued ZTE, Samsung, and Google, and apparently got those three companies to settle before also suing Apple.

  • KIP CR (= Crossroads) P1 has sued a number of companies including Huawei and Oracle. That entity even challenged the constitutionality of PTAB IPRs, but the Supreme Court denied that cert petition.

This problem is undoubtedly a whole lot bigger and more severe--and, therefore, more harmful to industry and consumers--than conventional "patent trolling." It will be interesting to see what else comes to light in the course of this litigation. Finally, here's the complaint:

19-10-21 Intel Antitrust Co... by Florian Mueller on Scribd

Share with other professionals via LinkedIn: