It has come to our attention that supporters of the Patent Trial and Appeal Board (PTAB) are claiming that the facts and figures presented in various IPWatchdog.com articles are erroneous. That is not the case. In fact, the criticism of some borders on outright misrepresentation.
As the facts laid out in this article will show, the PTAB is substantially more likely to find patents to be defective than a Federal District Court, and in fact have found numerous patents to be defective even after patentability has been confirmed in federal courts. This reality presents a significant problem for patent owners, and should be a serious concern for anyone at all concerned with separation of powers.
Let’s begin with a summary of key statistics and major points:
- 69% of cases reaching a final written decision by the PTAB have all claims invalidated.
- 168 out of 220 patents found valid in Federal Court (76.5%) were found defective by PTAB.
- 58 patents found valid in Federal Court were found defective by PTAB despite PTAB considering the same grounds of invalidity as the Federal Court.
- 82.5% of patents reviewed by PTAB in a final written decision are found defective.
- A settlement is erroneously characterized as a win for the patent owner.
- While claims don’t change if patent owner disclaims, they are unenforceable, which occurs in 5% of all petitions.
Lex Machina Data 9/16/2012 through 12/21/2017
Lex Machina™ is a LexisNexis® Company that mines litigation data, including data relating to post grant challenges at the PTAB. The data mined by Lex Machina™ reveals insights that have been unavailable. Thanks to this comprehensive data set and research tools, we have a clear picture of what is actually happening at the PTAB.
- Only 17% of all petitions are denied institution.
- 16% of all petitions result in all claims invalidated in a final written decision.
- 3% of all petitions result in at least some claims invalidated in a final written decision.
- 23% of all petitions are settled (13% before institution & 10% after institution).
- 5% of all petitions result in the patent owner disclaiming challenged claims.
According to Lex Machina™ data, 1833 cases have reached a final written decision as of December 21st, 2017. Of those cases that have reached a final written decision on the merits, only 17.2% (316 out of 1833) have had all instituted claims upheld and determined patentable while 69.2% (1268 out of 1833) of cases reaching a final written decision on the merits have had all instituted claims invalidated and determined unpatentable. There has been a mixed result in 13.3% (243 out of 1833). This means that in 82.5% of petitions that reach a final written decision, one or more claims are determined invalid and unpatentable.
It is worth noting that these statistics are consistent with USPTO figures released in October 2017 including data through September 30, 2017 which showed that 65% of final written decisions result in all claims invalid and unpatentable, 16% of final written decisions with at least some claims invalid and unpatentable and 19% of final written decisions where all claims are confirmed as valid and patentable.
An 82.5% defect rate in issued patents should be considered extremely alarming for a variety of reasons. First, and most obvious, an 82.5% defect rate on its face is unacceptable and should signal very serious problems either with examination quality or with an aggressively overactive PTAB. Instead, the United States Patent and Trademark Office (USPTO) claims that its examiners are correct in 92.4% to 96.5% of actions taken (see 2017 annual report, page 53), but also continues to support the actions and quality of the PTAB. An 82.5% defect rate in contested patents seems completely incongruous with any rational definition of high-quality examination or patent applications.
Second, the patents that are being challenged at the PTAB are among the most thoroughly vetted patents. “Since 80% of the IPRs are already in district court, these are patents the owners have already assessed, looking at whether infringement can be proven, whether the defenses against invalidity can be defended,” Chief Judge Paul Michel (CAFC, ret.) explained at a conference on Capitol Hill in May 2017. “These are presumably the strongest patents… a two-thirds invalidation rate can’t be right for these pre-vetted patents.” Of course, the invalidation rate is higher than two-thirds. According to Lex Machina data, all claims are lost in an astonishing 69.2% of cases reaching a final written decision.
Third, with 69.2% of challenged patents reaching a final written decision where all challenged claims are lost, and an overall defect rate of 82.5% of patents reaching a final written decision, what must be the true quality of issued patents? The patents worth fighting for are the most commercially valuable patents issued by the USPTO. If there is such a high error rate for the patents that matter the most to patent owners and technology companies, the error rate would have to be astonishingly high when considering those un-vetted, not commercially relevant patents.
Of course, the alternative conclusion has to be that the PTAB is aggressively overactive and is invalidating patent claims that deserve to be adjudicated valid. As the next section shows, that is almost certainly the case given the extraordinary difference between invalidation rates at the PTAB and in federal district court.
Docket Navigator® Data
While Lex Machina™ data focuses on petitions, with Docket Navigator® it is possible to focus on patents. According to Docket Navigator® data there are 3,055 patents for which no final decision has been reached. There have been 1,582 patents that have been the subject of a final written decision by the PTAB and at least one claim has been found defective in 1,343 patents. That means the PTAB is finding that 85% of issued patents they review are defective to one extent or another (i.e., 1,343 out of 1,582).
Docket Navigator® also makes it possible to focus on those patents where there has been both a Federal District Court ruling on validity and a PTAB final written decision relating to validity.
There have been 220 patents that were found to be valid in various Federal District Courts that were also reviewed by the PTAB. Of those, 52 patents were determined to have valid claims by the PTAB whereas 168 patents had claims, which were determined invalid by the PTAB. This means that 76.4% of those 220 patents found valid in Federal District Court were found defective by the PTAB.
A claim has been made by some that using Docket Navigator® data for this purpose yields unreliable statistics because in Federal District Court challenges to validity can be made on the basis of 35 U.S.C. § 101, § 102, § 103 and § 112, while at the PTAB only challenges based on § 102 and § 103 maybe be made in an inter partes review (IPR). That criticism is disingenuous to say the least. With more opportunities and theories to invalidate patent claims in Federal District Court, one would expect the invalidity rate to be higher, not lower, in Federal District Court. Only being able to apply § 102 and § 103 (and not § 101 and § 112), the PTAB still found 76.4% of the 220 patents found valid in Federal District Court to be defective.
Notwithstanding, critics have argued that this is not an apples-to-apples comparison because it is possible for claims to have been upheld in Federal District Court without a District Court Judge ever considering § 102 or § 103. Apparently, the argument goes like this: The patent owner may prevail on § 101 and/or on § 112, and that might provoke the defendants to settle, so we will never know whether the claims would have survived challenge under § 102 and/or § 103 in Federal District Court. Such an argument is flawed on its face. Apparently the critics would have us believe that defendants thought they had winning arguments based on prior art but settled anyway in Federal District Court after receiving an adverse decision on § 101 or § 112? Anyone familiar with patent litigation would easily understand that in the pervasively anti-patent environment of the moment that a defendant would not settle in Federal District Court despite believing they had winning arguments on validity. To suggest otherwise is a fantasy; it is simply counter-factual.
Regardless of the twisted logic that suggests defendants settle in Federal District Court despite the fact that they believe they would prevail on § 102 and/or § 103, we know with absolute certainty that there are numerous examples where there is a stark difference between what has occurred in Federal District Court compared to what has happened at the PTAB. Of those 168 patents found valid in Federal District Court but later invalidated by the PTAB we found an interesting subset of 58 patents. In this subset the Federal District Court specifically addressed validity challenges under § 102 and/or § 103, confirming that the claims were valid. Notwithstanding the fact that these patents had the challenged claims adjudicated to be valid in Federal District Court, the PTAB found the claims of those 58 patents to be invalid on the same grounds previously litigated.
Far more telling than any statistics are the specific instances where the PTAB has effectively overruled the rulings of Article III federal courts. Consider the following:
- Uniloc’s U.S. Patent No. 5,490,216 had been asserted in 65 cases filed in U.S. district court going back to September 2003, without claims ever being found invalid. After adopting a previous claim construction from the District Court the PTAB still found the challenged claims invalid.
- VirnetX patents had been awarded nearly $1.1 billion in damages, not including sums paid to settle disputes prior to a final decision in Federal District Court. All of VirnetX patents survived every challenge of invalidity in Federal District Court. Furthermore, the United States Court of Appeals for the Federal Circuit itself had previously confirmed the validity of the ‘135 and ‘151 patents. Notwithstanding, after multiple IPR petitions at the PTAB, the PTAB invalidated claims to the ‘151 patent, the ‘135 patent, the ‘504 patent and the ‘211 patents.
- The International Trade Commission issued a final determination finding a Section 337 violation by Arista relating to infringement of Cisco’s ‘577 and ‘668 patents. Not finding any of the invalidity defenses persuasive, and infringement ongoing, the ITC issued a limited exclusion order and cease-and-desist order against Arista. The PTAB declared claims of both the ‘557 and ‘668 patents invalid.
- Imperium IP Holdings’ U.S. Patent No. 7,092,029 survived 35 U.S.C. § 103 challenges for obviousness made by Samsung in a district court trial where enhanced damages were awarded by the court both for willful infringement as well as the presentation of demonstrably false testimony at trial. A jury decided that Samsung had not demonstrated obviousness of the ‘029 patent. Samsung then took its Section 103 challenge to the PTAB and the PTAB invalidated the patent on obviousness.
- Eight patents owned by Wisconsin-based Ultratec were twice upheld as not invalid in district court, once on summary judgment after a challenge on 35 U.S.C. § 102 grounds for anticipation and another time on Section 103 obviousness grounds in the jury verdict entered in the case. The defendants in that case, Sorenson Communications and CaptionCall, then moved on to the PTAB and had all challenged claims from the eight Ultratec patents asserted at district court invalid under Sections 102 and 103.
- Jazz Pharmaceuticals asserted 18 patents in district court in response to abbreviated new drug applications (ANDAs) filed with the U.S. Food and Drug Administration (FDA) for generic versions of Jazz’s Xyrem narcolepsy drug. The ANDA, filed by Amneal, included a Paragraph IV certification alleging invalidity of Jazz’s patents on multiple statutory grounds. One defendant in the case, Wockhardt Bio, entered into a consent judgment where they agreed that the generic drug would infringe Jazz’s patents and stipulated to an injunction against producing the generic. Wockhardt, Amneal and other defendants in the case then turned to the PTAB to invalidate six Jazz patents after defenses of patent invalidity didn’t work at the district court.
- Twice, Novartis AG’s U.S. Patent No. 6335031 was upheld as not invalid in district court after defendant Noven Pharmaceuticals challenged the patent’s validity on grounds of obviousness, once in a trial opinion from August 2015 and then again in a final judgment issuing from the court a few weeks later. Noven and fellow pharmaceutical firm Mylan then teamed up at the PTAB to have the ‘031 patent deveined of seven of the patent’s original 20 claims after the PTAB invalidated them for obviousness.
- Four patents owned by American firms Paice and The Abell Foundation were asserted in a case against Korean automakers Hyundai and Kia where those giant auto manufacturers were found guilty of willfully infringing the patents-in-suit. In a jury verdict issued in October 2015, none of Paice’s asserted patents were found invalid in light of Hyundai and Kia’s anticipation and obviousness arguments. The reasonable royalty awarded in the case was nearly $29 million, and that was prior to any enhanced damages for willfulness. And then Hyundai and Kia got by with a lot of help from their friends, including Ford and Volkswagen, at the PTAB, where those four patents were invalidated as obvious. One of these patents, U.S. Patent No. 7237634, had 26 PTAB trials petitioned against it, nearly 9 times the number of district court cases where that patent was asserted. In 13 PTAB cases, including four petitioned by Hyundai and Kia, the patent owner settled prior to institution. In the other 13 PTAB trials, each of which was petitioned by Ford, the PTAB found all challenged claims invalid on obviousness grounds.
- Two patents, U.S. Patent Nos. 8023580 and 8457228, were found valid and infringed in a jury verdict entered in district court in February 2015 against Samsung; the verdict also awarded $15.7 million in reasonable royalty damages to patent owner Rembrandt Wireless. Samsung had raised invalidity of the ‘580 and ‘228 patents on multiple statutes as affirmative defenses in the district court case. Although these patents were only asserted in one district court case, seven IPRs were filed against the ‘228 patent and another six petitioned to challenge the ‘580 patent, and Samsung was the petitioner in each of those cases. In three separate IPRs, the ‘228 patent was found invalid as obvious. Two of the six IPRs petitioned against the ‘580 patent also invalidated claims on grounds of obviousness.
Obviously, the PTAB is substantially more likely to find patents to be defective than a Federal District Court. This reality is a significant problem for patent owners, and should be a serious concern for anyone at all concerned with separation of powers. Even after prevailing in Federal District Court, and even after prevailing at the Federal Circuit as VirnetX had done, the PTAB can still invalidate claims already adjudicated as valid by an Article III tribunal. If patents are to be considered any kind of property right (as the statute says) title must at some point quiet, and an Article I administrative tribunal simply cannot have the power to overrule an Article III tribunal.
If patent owners cannot have full faith and confidence in the patent granted by the Federal Government, and they similarly cannot have full faith and confidence in a final adjudication by the federal courts, how can they be expected to invest the millions, and sometimes billions, required to bring technology to the marketplace?
Rebutting USPTO Statistics
Some have criticized IPWatchdog statistics by pointing to a USPTO power point presentation that is purported to show that 58% of patents survive IPR. That figure should cause anyone with vague familiarity with the IPR process and the PTAB to pause because that does not comport with practice realities. Upon a closer review, what is being cited is a USPTO power point presentation that discloses the results of a Multiple Petition Study conducted by the USPTO. According to the USPTO, 16% of petitions are filed after a decision on institution. First and foremost, this Multiple Petition Study confirms that petitioners are in fact ganging up on patent owners in a meaningful percentage of cases, which has been denied as fantasy by supporters of the PTAB.
Second, the Multiple Petition Study allegedly concludes that 58% of patents survive unchanged at the end of AIA proceedings, and that 69% of petitions result in a patent being unchanged. In order to get to numbers that high one must include in the “unchanged” category all settlements (23%), all denials of institution (17%), patent owners who disclaim (but do not change) the claims (3.5%), and those petitions that are procedurally dismissed before or after institution (9%). Based on Lex Machina™ data, including all of those categories of claims that remain “unchanged” results in 52.5% of petitions.
Of course, claiming that 52.5% of petitions result in patent claims unchanged fails to tell the entire story, and actually creates a false narrative that is divorced from reality. The PTAB does not include claims not instituted in a final written decision, which means the 17% of petitions not instituted do not end the story for the patent owner. In a number of situations patent owners have faced repeated challenges on the same claims after prevailing at institution. Similarly, when institution is granted the PTAB does not include those claims not instituted in a final written decision, which means no estoppel applies when there is no institution, but no estoppel also applies when there is institution but not to those claims the PTAB does not review.
Second, it is important to acknowledge that a patent owner disclaiming challenged claims is not a win for the patent owner because the patent owner has effectively invalidated those claims by giving them up, even if they technically remain unchanged. A disclaimer makes the claims unenforceable, which is the practical equivalent to invalidation.
One of the biggest mistakes the USPTO and proponents of the PTAB make is with respect to characterization of a settlement, which they characterize as a win for the patent owner because claims remain unchanged. Simply stated, a settlement is no win for the patent owner.
For example, according to Unified Patents (via their website), they only settle challenges with a no-cost license (i.e., they say they never pay to settle). Unified Patents has many member clients that gain the benefit of any such no-cost license settlement, which means the patent owner has capitulated and settled only to put an end to the costs associated with fighting. The patent owner receives nothing, and worse, there has been an initial determination that the challenged claims were invalid. So while the claims technically survive unchanged, a heavy cloud hangs over those claims the PTAB determined as likely invalid. Further, to the extent those claims remain at all viable in the real world, they have been effectively licensed to numerous companies having received a no-cost license.
Unreliable USPTO Statistics – Zond
Finally, it is also critical to understand that PTAB statistics provided by the USPTO are of questionable veracity.
The best illustration that demonstrates the highly unreliable nature of USPTO statistics relative to PTAB proceedings come from the petitions filed against patent owner Zond. Before the AIA they owned 371 claims. By the end of the 125 IPRs filed against them (yes, 125 separate challenges) they owned 0 claims (i.e., all claims lost as invalid), which obviously corresponds to a 0% survival.
Every claim Zond owned ended up invalidated, yet according to the USPTO the institution rate against Zond was 88.6%. How can Zond have lost 100% of their claims and the statistics reported by the USPTO correspond with an 88.6% institution rate?
Due to the multiple challenges filed against Zond, 1,220 claims were instituted out of 1,377 claims challenged. These are not misprints or mistakes. Zond had 371 unique patent claims, but because of the multiple IPR petitions filed against all of their patents those 371 claims were challenged a total of 1,377 times, but instituted only 1,220 times. Thus, the number of claims that were instituted is more than three times the number of unique claims Zond owned. In other words, the institution rate shouldn’t have been 88.6%, it should have been closer to 329%!
The Zond example is dramatic evidence of questionable PTAB statistics provided by the USPTO. This is why it is essential to look to non-biased, independent, outside sources of data for a complete picture. The highly unreliable nature of USPTO statistics is exactly why Lex Machina™ and Docket Navigator® data must be considered in order to get a real picture.