Lynk Labs: How the Least-Vetted Documents Destroy Issued Patents

by Dennis Crouch

The Supreme Court is currently considering whether to grant certiorari in Lynk Labs, Inc. v. Samsung Electronics Co., No. 25-308, a case that could reshape the evidentiary foundation of inter partes review (IPR) proceedings. The question presented asks whether patent applications that became publicly accessible only after the challenged patent’s critical date qualify as “prior art consisting of . . . printed publications” within the meaning of 35 U.S.C. § 311(b). In its January 2025 decision, the Federal Circuit held that published patent applications may serve as prior art in IPR proceedings as of their filing date under § 102(a)(2), even if those applications were confidential at the time the challenged patent was filed. Lynk Labs, Inc. v. Samsung Electronics Co., 125 F.4th 1120 (Fed. Cir. 2025). The petition has attracted substantial amicus support. I had begun drafting my own amicus brief in support of the petition but did not complete that project. Instead, I offer some of my thoughts here.

At its core, the case concerns what I have called the “secret springing prior art” problem: a reference that was secretly on file at the USPTO at the time a patent application was filed, but which only becomes publicly accessible when it is later published. See Dennis Crouch, Secret Springing Prior Art and Inter Partes Review, Patently-O (Oct. 2024). The Martin reference at issue in Lynk Labs was filed in April 2003 but did not publish until October 2004, well after Lynk’s February 2004 priority date. Under 35 U.S.C. § 102(a)(2), Martin qualifies as prior art because it was on file before Lynk’s application, even though it was entirely non-public at that time. The question is whether such a reference can form the basis of an IPR challenge, given that § 311(b) limits IPR to challenges based “only on the basis of prior art consisting of patents or printed publications.” The Federal Circuit held that it can, reasoning that “printed publication” is “temporally agnostic” and that the prior art status of a reference is determined separately from its status as a printed publication. I find that reading difficult see – particularly with our long tradition that a printed publication is defined by its date of publication.

The Federal Circuit’s error lies in conflating two distinct inquiries: whether a reference qualifies as prior art under § 102(a)(2) and whether it qualifies as a printed publication under § 311(b). The result is that abandoned, unexamined applications are backdated to destroy issued patents, privileging the least-vetted references over genuine publications that were actually accessible to the public before the critical date.

A Closed Evidentiary Universe. Section 311(b) limits IPR to challenges brought “only on the basis of prior art consisting of patents or printed publications.” 35 U.S.C. § 311(b). Congress chose fixed, public, documentary objects whose status as prior art is ascertainable and temporally verifiable. These are artifacts whose legal identity arises from a public act: issuance for patents and publication for printed works.

The Federal Circuit’s decision distorts that limitation through what may fairly be called temporal gymnastics, stretching time to treat the future act of publication as if it had already occurred. By importing the legal fiction of “secret springing prior art” from § 102(e)/102(a)(2) into § 311(b)’s closed evidentiary universe, the court effectively transformed an unexamined, non-public filing into a retroactive “printed publication,” collapsing the distinction between fact of publication and fiction of priority. See Dennis Crouch, Publications Before Publishing and the Federal Circuit’s Temporal Gymnastics, Patently-O (Jan. 2025).

The Temporal Logic of Publication. A reference qualifies as a “printed publication” only when it is publicly accessible before the critical date. In re Klopfenstein, 380 F.3d 1345, 1348 (Fed. Cir. 2004); In re Hall, 781 F.2d 897, 899 (Fed. Cir. 1986). That requirement reflects more than a century of patent practice distinguishing the fact of publication from the legal fiction of priority. The “secret springing prior art” doctrine originated in Alexander Milburn Co. v. Davis-Bournonville Co., 270 U.S. 390 (1926), and serves a specific function: preventing the issuance of overlapping patents to applicants filing close in time. But that doctrine answers only the question of prior art status.  It does not transform the document’s nature. A secretly-filed application that later publishes may be deemed prior art as of its filing date, but it does not thereby become a “printed publication” as of that date. Section 311(b) asks a different question: not merely whether a reference is prior art, but whether it is prior art consisting of a patent or printed publication.

Section 311(b) is object-based and time-fixed. It presupposes that the public act converting a document into a printed publication occurred before the patent’s critical date. The Federal Circuit’s contrary view, that “printed publication” is “temporally agnostic,” erases the temporal boundary that gives the phrase meaning. Publication is the birth of a printed publication, not an event that can be retroactively imputed to the past.

Historical Continuity. The formulation used in the 2011 AIA statute was lifted directly from the same language found in the original 1980 reexamination statute, 35 U.S.C. § 301, limiting reexamination to “prior art consisting of patents or printed publications.” At that time, no U.S. patent applications were published at all. Every application remained secret until issuance. Thus, “printed publication” could not possibly have included unpublished filings. Two decades later, the American Inventors Protection Act of 1999 introduced 18-month publication, 35 U.S.C. § 122(b)(1)(A), and regular U.S. pre-grant publications began in 2001. Yet when Congress created IPR in 2011, it reused the exact § 301 phrase: “patents or printed publications.” Nothing in the legislative record suggests Congress intended that familiar language to sweep in retroactive backdating of published applications. Repeating identical statutory language signals continuity, not expansion.

Information Quality Concerns. Congress’s choice to confine IPR evidence to patents and printed publications also functions as an information-quality filter. Issued patents and genuine publications have undergone at least one form of substantive or editorial review. Unexamined and abandoned applications have not. As Professor Janet Freilich has shown, “information quality” is a recurring problem in patent examination: the Patent Office and the public often rely on incomplete, speculative, or erroneous disclosures that distort the perceived state of the art. See Janet Freilich, Ignoring Information Quality, 89 Fordham L. Rev. 2113 (2021). Empirical work confirms that abandoned or unexamined applications are particularly unreliable as technical teaching. See Lidiya Mishchenko, Thank You for Not Publishing (Unexamined Patent Applications), 47 BYU L. Rev. 1563 (2022) (finding that such applications often lack enabling detail and can mislead later reliance).

The Lynk Labs decision creates perverse incentives by treating abandoned published applications as superior to other printed publications through backdating. It grants the least-vetted documents the greatest effect, allowing filings that failed to meet the Patent Office’s own standards to destroy patents that did. This outcome contradicts the AIA’s animating goal: a transparent, quality-based patent system tied to what the public actually knew, not what it might later be deemed to have known.

Doctrinal Coherence. The Federal Circuit has itself recognized that § 311(b) sets a bright-line evidentiary rule: IPRs may rest only on patents or printed publications. Qualcomm Inc. v. Apple Inc., 24 F.4th 1367, 1376 (Fed. Cir. 2022) (holding that applicant-admitted prior art cannot form the basis of an IPR ground). That principle was extended by the USPTO’s own 2025 memorandum forbidding reliance on “general knowledge” or applicant-admitted prior art as claim-element substitutes. These decisions collectively reflect the same structural insight: § 311(b) cabins IPR to public, documentary, and temporally fixed evidence. Lynk Labs departs from that settled understanding. Its “temporally agnostic” construction destabilizes a decade of coherent doctrine and converts a closed evidentiary system into an open-ended one.

The Amicus Landscape. Several amicus briefs filed in support of the petition deserve mention. Former USPTO Director David Kappos, former Federal Circuit Chief Judge Paul Michel, former House Judiciary Chairman Lamar Smith, and IP practitioner Philip Johnson filed jointly, emphasizing their direct involvement in drafting and negotiating the AIA. Their brief makes the legislative insider argument: that the “two-window compromise” deliberately constrained IPRs to a narrower scope than PGRs and district court litigation, with § 311(b)’s language serving as one of the “effective protections” built into the compromise. The Alliance of U.S. Startups and Inventors for Jobs (USIJ) filed a brief highlighting how the PTAB’s expansive jurisdiction destabilizes the innovation economy, noting that the PTAB has cancelled, in whole or in part, over 80% of patents reaching the final written decision stage.  Of course, we recognize that things have changed at the USPTO under Director Squires and now are increasingly favoring the patentee in IPR proceedings.

Professor Tim Hsieh filed an amicus brief arguing that the Federal Circuit’s decision conflicts with Loper Bright Enterprises v. Raimondo, 603 U.S. 369 (2024), and effectively resurrects Chevron deference in disguise. See Tim Hsieh, Guest Post: The Supreme Court Should Clarify How to Apply Loper Bright in the Patent Law Case of Lynk Labs, Inc. v. Samsung Co. Ltd., Patently-O (Dec. 2025). Hsieh contends that the Federal Circuit parroted the USPTO’s policy-driven interpretation rather than exercising independent judicial judgment, treating agency policy as determinative of statutory meaning in a manner Loper Bright forbids.

The intellectual property scholars’ brief, spearheaded by Professor Ted Sichelman of the University of San Diego (joined by Professors Emily Morris, Joshua Kresh, and Mark Schultz), takes a textualist approach that warrants more extended discussion. Their brief argues that applying established textualist principles makes the errors in the Federal Circuit’s interpretation “crystal clear.” The scholars emphasize that the language “application(s) for patent” or “patent application(s)” appears in numerous other sections of the Patent Act, including those related to prior art and validity such as §§ 100 and 102, but not in § 311(b). Under the canon expressio unius est exclusio alterius, this omission is meaningful. Congress plainly did not pursue any policy of expansive prior art availability to its logical endpoint in IPR; if ease of handling were the touchstone, then any prior-art reference documented in writing would be fair game, including written proof of prior use or prior sales.

The scholars’ brief makes a particularly incisive point about statutory surplusage. If the Federal Circuit’s interpretation were correct, the separate inclusion of “patents” in § 311(b) would have no function. Just like published patent applications, issued patents are also necessarily printed publications. So if “printed publications” allows published patent applications to serve as prior art as of their filing date under § 102(e)(1), then it must do the same for patents under § 102(e)(2). The separate, express inclusion of “patents” alongside “printed publications” would add nothing.

Next Steps: The Supreme Court has indicated some interest in the petition – requesting Samsung to file a responsive brief.  Samsung has requested an extension and so their brief is due in early January 2026 along with amici supporting the broader reading of the statute adopted by the Federal Circuit.