The Federal Circuit’s recent decision in Google LLC v. IPA Technologies, Inc., Appeals 2021-1179, -1180, and -1185 (Fed. Cir. May 19, 2022), offers three instructive reminders. First, a publication disqualified from consideration as prior art during prosecution may be resurrected as prior art during an AIA trial. Second, the burden of producing evidence is not static, but rather one that shifts among trial participants. Third, when presented with a “highly relevant evidentiary conflict”—like conflicting fact-witness testimony—the Patent Trial and Appeal Board must resolve it “and make appropriate findings of fact.” Google, Slip Op. at 11. When it doesn’t do this, it’s getting the case back.
Resurrecting Prior Art Isn’t Easy
A long time ago, two patents issued after the two named inventors (Messrs. Martin and Cheyer) persuaded the Patent Office that an earlier publication they authored with a third person (Dr. Moran) was not prior art. The publication described some of the underlying technology embodied in the claims of the patents. Unsurprisingly, the Patent Office’s examiner rejected the claims. After all, the examiner had no evidence as to who among the three conceived the underlying technology. If it was Dr. Moran, the rejection would remain sound because, then, the inventive entity of the patents (two people) would be different from the three authors of the publication. But if one or both of the named inventors was responsible, then the rejection could not be sustained. See generally, In re Katz, 687 F.2d 450, 454–56 (CCPA 1982) (explaining how an applicant may disqualify a co-authored publication from consideration as prior art under 35 USC § 102(a)).
The inventors submitted declarations summarily concluding that each was a co-inventor of the claimed invention and that Dr. Moran was “not a co-inventor.” The inventors, however, offered no evidence that they understood the legal concept of inventorship. More significant, the declarations did not explain who among the three was responsible for having invented the underlying technology reported in the publication and on which the examiner relied to reject the claims. See In re DeBaun, 687 F.2d 459, 462 (Fed. Cir. 1982) (concluding that declarations are insufficient where “they fail to allege facts showing the necessary diligence and/or reduction to practice of the invention now claimed”); see also, In re Katz, 687 F.2d at 455 (“What is required is a reasonable showing supporting the basis for the applicant’s position.”). Incredibly, however, the examiner accepted these facially deficient declarations, withdrew the rejection, and eventually allowed the patents to issue.
Years later, IPA Technologies asserted the patents in federal court against Google, Microsoft Corporation, and Amazon.com, Inc. In response, Google successfully petitioned the Board to institute inter partes review of the patents, arguing the previously-disqualified publication rendered the patents invalid, notwithstanding the inventors’ declarations. Google also produced Dr. Moran, who testified that he—not the named inventors—was responsible for the relevant underlying technology in the publication. PTAB Institution Decision, Case IPR2019-00728, Paper 11, at 12 (PTAB Sept. 4, 2019). The declarations the named inventors presented to disqualify the publication during prosecution were suddenly neutralized. See id. at 16–17 (agreeing with Google “that the inventor declarations were insufficient to disqualify [the publication] as prior art,” and “the Examiner should not have disqualified [the publication] … based on [those] declarations”). A trial before the Board ensued.
Assessing whether the publication is prior art to the patents is, according to the court, like an inventorship assessment—a question of law based on underlying facts. Google, Slip Op. at 4. The court identified three relevant factual inquiries. First, is an assessment of what parts of the publication are relevant to reject the patent claims. Next, is a determination of who conceived each of those parts. It may be that some of those parts were conceived by the individual(s) not also named as inventors on the patent. In that situation, the Board also needs to assess whether the non-inventor’s contribution—relative to the rest of the relevant disclosure in the publication—is sufficiently significant to render him/her a joint inventor of the relevant disclosure in the publication. Id. at 5 (citing authorities). But, before the Board, who has the burden of proving this?
The Burden of Production Is a Shifting Burden
The court noted that the burden of proof includes two concepts. One concept is the burden of persuasion, an unshifting burden the AIA trial petitioner must satisfy to prove unpatentability. The other concept is the burden of production, a shifting burden to present evidence. Who among the trial participants has this burden at any given time depends on the particular issue, and it may be met by presenting evidence or argument. Id. at 5–6.
The court explained that Google, as petitioner, had the burden of persuasion, and saw no error with the Board requiring Google to produce evidence that Dr. Moran made an inventive contribution to the portions of the publication relied on and relevant to establishing the patent claims are obvious. Id. at 7. That’s not an easy task. Witnesses may not willingly cooperate and even if they do, over time, memories often fade. Here, Google produced evidence (Dr. Moran’s testimony) and argument with its petition, consistent with the burden of production it had to initially carry. After institution and during the trial, IPA Technologies subpoenaed Messrs. Martin and Cheyer and subsequently obtained their direct testimony recounting the activities relating to the publication, all in an effort to rebut Dr. Moran’s testimony. PTAB Final Written Decision, Case IPR2019-00728, Paper, at 14 (PTAB Sept. 2, 2020). But instead of assessing Dr. Moran’s testimony (relative to the direct testimony of the named inventors), the Board essentially dismissed it, concluding it was insufficiently corroborated to qualify the publication as prior art, entering judgment against Google. Id. at 23–25. The court rejected the Board’s conclusion as unsupported by the factual record, which contains “more than adequate corroboration.” Google, Slip Op. at 10 (listing the corroborating evidence).
The Board Must Resolve Conflicting Testimony
All three individuals testified during the trial, offering apparently conflicting testimony about whether Dr. Moran played a sufficiently significant role in conceiving the underlying technology. Id. at 7–8, n.4. The Board, charged with fact-finding, found all three witnesses credible, but did not “resolve this highly relevant evidentiary conflict and make appropriate findings of fact.” Id. at 10–11. The court therefore vacated the Board’s decision and remanded for appropriate resolution of whose testimony is more credible.
This case presents a unique set of facts and challenges. The Board rarely administers cases presenting conflicting fact-witness testimony. Further, the Board rarely—if ever—observes any witnesses presenting their testimony, and rarely poses questions to those witnesses. These are classical ways district courts assess witness credibility. Yet, the Board is tasked here with assessing which credible witness is more credible. On remand, will it entertain live testimony to reassess credibility? Its forthcoming decision may pivot on that factual finding, which the court, should another appeal ensue, will review deferentially.