Published Patent Applications Are Prior Art as of the Filing Date, Not the Publication Date
By Joshua Rich --
Lynk Labs raises a simple question of statutory interpretation with surprisingly important ramifications: in inter partes review, is a published patent application considered prior art as of the date of filing or the date of publication? The former answer has been how the law has long been applied and would comport with Federal court practice, but Lynk argued that Congress had chosen the latter approach by keeping the language that had been in the statute for decades before IPRs existed and not expressly including patent applications as a separate category of available art. The Federal Circuit found that the statutory language governed only whether a reference could constitute prior art, not when it would. As a result, the Court rejected Lynk's approach and decided that patent applications, like patents, are treated as prior art as of the filing date.
The issue arose out of Samsung's IPR challenge to Lynk's U.S. Patent No. 10,687,400. The '400 Patent relates to alternating current ("AC")-driven light emitting diodes ("LEDs"), LED circuits, and AC drive circuits and methods. Basically, it covers the type of lighting now used in many Christmas tree lights: "an LED circuit array comprising an LED circuit comprising a plurality of LEDs connected in series" and a number of other elements, "wherein a forward voltage of the LEDs of the LED circuit array matches the rectified input AC voltage output" of a connected driver.[i]
Samsung brought its IPR based on an argument that the claims were obvious in light of numerous combinations of prior art. Importantly, many of the combinations relied on U.S. Patent Application Publication No. 2004/0206970 (the "Martin Application"). The Martin Application was filed more than one year before the application that ripened into the '400 Patent, but published several months after the filing; it never issued as a patent. Given those circumstances, Lynk argued that the Martin Application was potential prior art as a "printed publication," not a published patent application (to the extent that is a separate category under the Patent Act) and would not qualify as prior art in this case because it was not printed until after the '400 Patent's critical date. The PTAB disagreed and found all of the relevant claims unpatentable.[ii]
Lynk appealed, arguing that IPRs are limited to "patents and printed publications,"[iii] and published patent applications could only be considered the latter. That is, unlike the pre-AIA § 102 that applied here, the scope of IPRs does not explicitly contemplate published patent applications in any context separate from the printed publications referenced in §§ 102(a) and (b). It follows, according to Lynk's reasoning, that published patent applications should be treated like all other printed publications for purposes of determining the prior art date, regardless of the provisions of § 102(e)(1) that relate specifically to published patent applications. Lynk argued that the use of the long-used term "printed publications" carried with it the "old soil" of prior uses and court opinions.
It is true that the terms "patents and printed publications" have long been used in patent law and post-grant review proceedings. Since 1980, there have been opportunities for post-grant review of patents: first ex parte reexamination (created by the 1980 Patent Act), then inter partes reexamination (created by the 1999 American Inventors Protection Act), and most recently post-grant review and inter partes review (created by the 2011 America Invents Act). Throughout that time, the identification of available prior art has remained the same: "patents and printed publications."[iv] But in truth, the parameters for what is available to challenge the validity of patents -- both in USPTO proceedings and in district court litigation -- have changed. Prior to the AIPA, the USPTO had maintained strict secrecy over patent applications so there were no published patent applications. Then, the AIPA ushered in the era of published patent applications and added the "secret prior art" provisions of § 102(e), establishing that a patent would be invalid if:
the invention was described in
(1) an application for patent, published under section 122(b), by another filed in the United States before the invention by the applicant for patent or
(2) a patent granted on an application for patent by another filed in the United States before the invention by the applicant for patent, except that an international application filed under the treaty defined in section 351(a) shall have the effects for the purposes of this subsection of an application filed in the United States only if the international application designated the United States and was published under Article 21(2) of such treaty in the English language
Those two changes under the AIPA set the stage for the Federal Circuit's decision in Lynk Labs.
Lynk had admitted that published patent applications constitute "printed publications," but argued that whether they are prior art should be considered under the paradigms of § 102(a) and (b) regardless of the language added by the AIPA. It argued that it would be anomalous to treat documents within the ambit of printed publications differently based on what type of printed publication they are. And it had on its side older court decisions that suggested that all printed publications should be treated as prior art only when they are published.
The Federal Circuit recognized that Lynk's position found some support in older case law, but reasoned otherwise based on the language of the entirety of § 102. It found the addition of § 102(e) dispositive because it specifically addressed the timing of prior art status of published patent applications. And responding to the anomaly Lynk pointed out, it noted another one if it applied the test urged by Lynk -- that published patent applications would not be treated the way set forth in § 102(e), but patents would be. In a bit of statutory interpretation legerdemain, the court suggested this was all just a plain language reading of §§ 311 and 102. A better approach would have been to use the canons of construction that suggest that the proper reading of a statute gives effect to all of the clauses and reads each of them in the context of the statute as a whole.
The Federal Circuit then took a confused, and confusing, trip through the cases that Lynk cited to suggest that published patent applications are only an unremarkable subset of printed publications. When Congress amends a statute but retains the same language as used before, it is presumed to be importing the prior usage of that language, which would suggest that "printed publications" would be defined consistently throughout the history of post-grant proceedings. But the court rejected that approach because published patent applications did not exist throughout that history. And it argued that printed publications could have different prior art dates based on the differential scenarios of § 102(a) and § 102(b). But that seems wrong -- the date of the printed publication does not change, only whether it is before or after the time point identified in the statute. That is, a journal article has a fixed date of publication -- just like a published patent application -- which may be before or after the date of invention or bar date for a claimed invention. In the end, the Federal Circuit concluded that "whether, and how, something that qualifies as a 'printed publication' is considered prior art depends on other statutory language" -- again defaulting to the language of § 102(e).
Remarkably, the Federal Circuit relegated to a footnote perhaps the best policy argument for interpreting the statute as it did. That is, in district court litigation, published patent applications and patents are both treated as prior art as of the date of filing.[v] Had the case come out the other way, accused infringers would have yet another reason to avoid IPRs because some of the patent applications considered prior art in district court proceedings would not be considered prior art in IPRs. Nothing about the references would be different, just the legal treatment of the reference between the two venues. Not only would such an approach increase the likelihood of gamesmanship in the choice of venue to resolve validity, it would also have created numerous other difficulties – Would estoppel apply to a published patent application like the Martin Application? Would the same approach apply in post-AIA cases, where the language of § 102 was altered slightly? And would a different approach apply in IPRs and ex parte reexamination?
In the end, regardless of the reasoning, the Federal Circuit reached the correct conclusion in Lynk Labs. A contrary decision would have made a mess of the interplay of IPRs and district court litigation, and would also have been inconsistent with the totality of § 102. And it would have called into question IPRs already decided and currently pending that have been decided or filed based on the presumption that the approach to published patent applications would not change suddenly after 25 years of clear precedent.
Lynk Labs, Inc. v. Samsung Electronics Co. (Fed. Cir. 2025)
Panel: Circuit Judges Lourie, Prost, and Stark
Opinion by Circuit Judge Prost
[i] For Taylor Swift fans, the timing of the Federal Circuit decision may draw a smirk because, as she sings in Lover, "We could leave the Christmas lights up 'til January."
[ii] Lynk Labs cancelled certain claims in response to the IPR filing, rather than having those claims invalidated.
[iii] 35 U.S.C. § 311(b).
[iv] The exception to this rule is post-grant review, for which any basis under 35 U.S.C. § 101 is available, including other forms of prior art.
[v] See pre-AIA 35 U.S.C. § 102(e) and current 35 U.S.C. § 102(d).
Is the application publication still treated as prior art as of its publication date when the application that is published is "not by another"?
Posted by: Chad D. Tillman | January 17, 2025 at 08:06 AM