The Federal Circuit regularly affirms PTAB judgments without issuing any explanatory opinion to justify the result. Although not found in the Rules of Appellate procedure, the court has created its own local rule allowing itself to "enter a judgment of affirmance without opinion." In a 2017 paper, I argued that these no-opinion affirmances violated both the spirit and letter of 35 U.S.C. 144, which requires the court to issue a "mandate and opinion" in cases appealed from the USPTO. Since that time, the Federal Circuit has continued its practice, issuing hundreds of no-opinion judgments. Throughout this time, dozens of losing parties have petitioned for en banc rehearing with the Federal Circuit or certiorari to the Supreme Court. Up to now, both courts have remained silent and have refused to address the issue.
A new pending petition raises the issue once again. Virentem Ventures v. Google (Supreme Court 2023). Virentem sued Google for patent infringement, and Google responded with a set of Inter Partes Review (IPR) petitions. The PTAB eventually sided with Google and invalidated the claims of all seven challenged patents. Virentem appealed; but the Federal Circuit affirmed the PTAB's judgement without opinion under its local Rule 36.
The new petition to the Supreme Court asks four related questions:
Does the Federal Circuit’s use of Rule 36 to affirm without opinion PTAB invalidity determinations that are challenged based on pure questions of law violate a patentee’s due process rights through arbitrary or disparately applied results?
Did the Federal Circuit’s use of Rule 36 to affirm without opinion PTAB invalidity determinations of Virentem’s patents violate its due process rights?
Did the PTAB’s adoption, and Federal Circuit’s summary affirmance, of broad constructions of Time Scale Modification and other claim terms over Virentem’s explicit narrowing definitions, violate the Federal Circuit’s own law and precedents on claim construction in such circumstances?
Does the Federal Circuit’s use of Rule 36 to affirm without opinion decisions from the PTAB violate the requirement of 35 U.S.C. § 144 that the Federal Circuit “shall issue to the Director its mandate and opinion”?
The Virentem patents relate to time-scale modification -- the speeding-up or slowing-down of media. You may remember Alvin, Simon, and Theodore -- the Chipmunks. That unintelligible high pitch arguably is not really time-scale modification because it is such a failure. Rather, TSM modern impliedly requires maintaining pitch and intelligibility. In this case though, the PTAB broadly interpreted the term to include any system that speeds-up or slows-down media. With that broad interpretation, the tribunal then was able to find prior art rendering the claims obvious. Virentem argued that its patents would be seen as valid under the narrower construction. The PTAB's response: If you wanted that limitation in the claim, you should have added it to the claim.
I've become somewhat callous toward patent eligibility jurisprudence and so was surprised when I read the Federal Circuit's decision in ADASA Inc. v. Avery Dennison Corp., 55 F.4th 900 (Fed. Cir. Dec 16, 2022). The case concluded that constructively dividing a fixed-length binary number into different sub-portions was not an abstract idea. Now, the accused infringer has asked the Supreme Court for review.
ADASA's US Patent 9,798,967 is directed toward an RFID chip "encoded with a unique object number." This object number as various blocks pre-allocated to identify the selling-company, product reference, and serial number as shown below. The serial number has a unique feature of being divided into a section of "most significant bits" (MSBs) and "least significant bits" (LSBs). This division between MSBs and LSBs help in the allocation of unique serial numbers in a distributed production system. The basic approach:
A product-line will be exclusively allocated a particular MSB and all possible accompanying LSBs. It will then enable RFID chips using the allocated MSB and then sequentially incrementing the LSB.
A separate product line might also be simultaneously enabling other other chips. To ensure no overlap in serial numbers, the second-line will be allocated a different MSB.
The result then is that we can guarantee that each chip has a unique serial number despite parallel production lines.
The claims do not appear to include any novel features other than this constructed division between bits in a binary number. (Claim 1, is reproduced below). In its petition, the adjudged infringer relies heavily on the old cases of Benson and Flook to argue that the setup here lacks eligibility.
Question presented:
Radio Frequency Identification Device (RFID) tags are encoded with lengthy serial numbers that uniquely identify particular items. The patent at issue in this case designates the leading bits in a binary serial number as “the most significant bits,” and directs that all serial numbers in an allocated block begin with the same “most significant bits.”
The question presented is whether that claim, by subdividing a serial number into “most significant bits” that are assigned such that they remain identical across RFID tags, constitutes patent-eligible subject matter under 35 U.S.C. § 101.
Chief Judge Moore wrote the opinion and concluded that the claims were not directed to an abstract idea but rather provides a novel data structure within a serial number. Here, the idea is that the patentee was able to create a new data field that was "not a mere mental process, but a hardware-based data structure focused on improvements to the technological process by which that data is encoded;" and an improvement with "important technological consequences."
An interesting feature is that the general idea presented by the patentee was already identified and discussed in the book RFID for Dummies. If Avery Dennison loses here, then the district court will hold a trial on anticipation. Still, the "for Dummies" label has strong rhetorical appeal -- should everything in that book title be considered an abstract idea? The petitioner writes:
“[A]bstract ideas are not patentable.” ... It is hard to imagine a more blatant transgression of that rule than the claim in this case, which sought a patent monopoly over the simple concept of treating one long serial number as the combination of two shorter numbers, and then [requiring] blocks of RFID tags all start with the same shorter number. At bottom, that claim is no different from a direction to mentally subdivide all telephone numbers into two component parts and then assign the same leading part to an allocated block of numbers (e.g., all telephone numbers in the District of Columbia start with 202). That may be a good idea (indeed, the concept in ADASA’s patent appears in the pages of RFID for Dummies), but it is no more patentable than the other good but abstract ideas that this Court has held unpatentable for more than 150 years.
Id. I'm confident that if the Supreme Court takes this case, it would even further expand eligibility doctrine.
Guest post by Jonathan Stroud. Stroud is General Counsel at Unified Patents – an organization often adverse to litigation-funded entities.[1] He is also an adjunct professor at American University Washington College of Law.
Patent assertion finance today is a multibillion-dollar business.[2] Virtually nonexistent in the patent space in the U.S. ten years ago—at least in part due to longstanding common law rules on champerty, maintenance,[3] and patent law’s relative high risk—today third-party litigation funding (TPLF)[4] undergirds about 30% of all patent litigation, by conservative estimates.[5] Insurance options are suddenly plentiful,[6] funders are expanding and multiplying,[7] and new deal commitments are on the rise.[8] This general trend is seen in the first chart below, adapted from a recent white paper by Korok Ray.[9]
That is in no small part due to it being the fastest-growing piece of the wider U.S. litigation finance boom of the past 20 years—as has been widely reported, private equity now undergirds huge swaths of U.S. bankruptcy, class action, trademark, securities, and tort litigation, to the tune of $50 to $100 billion in investments annually.[10] According to one of the biggest litigation funders, publicly traded Burford Capital—recently featured on 60 Minutes[11]—there was a 237% increase in overall litigation funding in the US between 2012 and 2018, a trend that, by all accounts, continues unabated.[12] Industry reports show new investments pouring fastest into patent infringement litigation; new deal commitments for TPLF saw an increase of 61%; and patent litigation accounted for 29% of all new commitments by TPLFs in 2021.[13] Recent trends are shown in the chart below, adapted from a Westfleet Advisors report. [14]
In terms of how TPLF is structured, deals are variegated, complex private agreements. But generally the funder will offer non-recourse funding (or funding that is “at risk”) upfront to cover expenses in exchange for being first in line to recoup all of that funding first (i.e., to be “paid back”) out of any recovery, and then to take some hefty percentage—often 60% or more of whatever is remaining, particularly in litigations deemed high-risk (like patent litigation), though there are no rules governing how much funders can ask for. (It generally amounts to more than 50% of the total settlement recovery, acknowledging, at least by basic math, that they are the primary beneficiary of the litigation.). Sometimes all fees are paid upfront by the funder (Fortress is known for this); some pay some continuing level of a fee/contingency split with firms to split risk; some pay the original patentholder upfront, though others think that disincentivizes them from robust ongoing participation; others make all recovery, for all parties in a waterfall, contingent upon settlement. Many start with and later add investors to ongoing funds and matters. Nearly all require oversight and consultation at all key decision points.
Patent TPLF funds generally promise roughly 20% internal rates of return to funders (IRR) year-over-year, or about a 2x to 2.5x return on investment over generally four- or five-year investment cycles, suggesting, at least at the pitch level, that these investments are lucrative for the funders.[15] The biggest (or at least most well-known) players—Magnetar Capital, Burford Capital, Fortress Investment Group, Omni Bridgeway, and Curiam Capital, to name just a few[16]—have funded patent cases for years, reporting in some cases that their existing funds were on pace to return 20% or more—less than some other investments tout, but still beating the market by a fair margin.[17]
At least, that’s as far as can be pieced together. What we do know comes mostly from self-reporting, industry reports, and journalists. That’s because current disclosure of litigation funding relies on a patchwork of state law, court rules, self-reporting, FOIA requests, leaks to journalists, and funding pitches. It’s true today that no one in the government (Federal or state, judicial, legislative, or executive) knows who is funding which litigations, whether they are as profitable as they claim to be, if they are being properly taxed, or even how they are generally structured. Disclosure is limited even for the two well-known, publicly traded litigation fund managers, Burford Capital and Omni Bridgeway; it is sparser still—and highly self-selective—for all the private funds involved. According to a recent Government Accountability Office (GAO) report on litigation funding (written at Congress’ behest), “[e]xperts GAO spoke with identified gaps in the availability of market data on third-party litigation financing, such as funders’ rates of return and the total amount of funding provided,” and noted that no government body is aware of who is funding these cases, who is influencing or controlling them, or what promises they are making to investors.[18] (It also notes litigation finance industry lobbying groups active today, and their membership.)
Disclosure remains sparse at least in part because the very wealthy private investors who fund litigation claims and then reap, they claim, windfall profits—some of them concededly foreign sovereign nation funds[19]—have fought hard to keep those agreements secret, even from judges asking for disclosure, much less from government officials, researchers, reporters, opposing parties, or the public. As such, the Federal District Court of Delaware has recently found itself at the center of this high-stakes debate about transparency and the purpose of the courts.
In April of 2021, the District of Delaware’s Chief Judge, Colm Connolly, issued two standing orders requiring litigants to, inter alia, disclose third-party litigation funding.[20] (The orders apply to all parties and litigation before his Court, not just parties to patent disputes, but do not extend, as yet, to the other sitting judges there.) The orders were neither ultra vires nor exceptional—The Federal Rules of Civil Procedure have been moving toward greater ownership transparency for years, the advisory committees have recommended that judges have the right to such disclosure and are considering further requirements,[21] and similar requirements in Federal District courts across the nation have been in place for years, in districts in, for example, California, Georgia, Iowa, Maryland, Michigan, Nevada, New Jersey, Ohio, and Texas (in the Western district).[22] But that trend toward disclosure had thus far largely avoided being raised and enforced in the few Federal districts where patent litigation primarily resides (though the California and Texas districts have long had rules requiring disclosures—ones that are often ignored by LLC PAEs).
