We have previously written about disputes regarding whether a prior art reference qualifies as a publication in AIA trials. In a recent decision, the Board ruled that a prior art reference did not qualify as a publication under §102(a). The case, Activision Blizzard v Acceleration Bay, rejected petitioner’s attempt to rely on a 1999 technical report allegedly published by University of California / San Diego (“UCSD”).
Key Takeaway: This case demonstrates that the PTAB will adhere to demanding requirements for petitioners to establish that a reference qualifies as a prior art publication. Petitioners should be prepared to provide detailed evidence showing a prior art reference was publicly accessible before the priority date of a challenged patent. Meanwhile, patent owners should take advantage of opportunities to challenge the public accessibility of references, particularly where the evidence and arguments in a petition are conclusory; such challenges can be effective in a preliminary response and may derail a case before trial is instituted.
Case Background: Petitioner filed an IPR petition challenging patent no. 6,829,634, which relates to a broadcast channel overlaying a point-to-point communications network. The PTAB instituted trial on two grounds, both based on a 1999 technical report authored by Lin et al. (“Lin”). Petitioner asserted that the Lin report was made publicly available via a website for the UCSD computer science department library, and included with the petition a declaration by a systems administrator at UCSD, along with a web printout showing a summary page for Lin as listed in the UCSD library and a copy of Lin.
Patent owner did not challenge the publication status of Lin in its preliminary response, instead asserting that the invention in the ’634 patent pre-dated Lin. But in its merits response, patent owner contested Lin’s qualification as a publication under §102(a), asserting that there was insufficient evidence to establish that Lin was publicly available in 1999. In reply, petitioner submitted additional arguments and evidence to support its position on the publication of Lin, including a rebuttal declaration by petitioner’s technical expert, a separate web page maintained by one of the co-authors of Lin, and a declaration by a library science expert.
Final Written Decision – Technical Report Not a Publication: Since both instituted grounds were based on Lin, the issue whether that reference qualified as a publication was potentially dispositive as to the outcome of the IPR. Referring to the oft-cited axiom that public accessibility is the “touchstone” of whether a document is printed publication, the Board applied the standard of public accessibility as set forth in the Federal Circuit’s Blue Calypso opinion: a reference is publicly accessible if it was disseminated or otherwise made available so that persons of ordinary skill in the art, exercising reasonable diligence, can locate it. The Board then analyzed each of the items of evidence cited by petitioner as evidence of public accessibility, and concluded that Lin was not publicly accessible. Accordingly, the Board found for patent owner on both grounds. Below, we discuss the key evidence and arguments on public accessibility considered by the PTAB.
UCSD Technical Reports Library/Declaration. Petitioner argued that the Lin report was publicly accessible because it was available via a library website maintained by the UCSD department of computer science. A systems administrator for the UCSD computer science department testified about the operation of the UCSD electronic technical reports library, the URLs for the library, and the department’s practice of receiving articles, assigning of a unique identifier to each report, and uploading to the site. Based on the summary page for the Lin report, as well as operating systems records, he stated that the report became available to the public in November 1999. However, the declaration did not otherwise indicate how technical reports are organized in the library or how a person of ordinary skill (“POSA”) would search for technical reports on the UCSD site. In particular, the declaration did not explain how the library website was indexed or searchable.
On cross-examination, the UCSD declarant testified that the library website has a search page providing a list of technical reports by author or by year. However, although there was an advanced search page that appeared to permit searching abstract fields by keyword, the declarant did not know how the search worked or how keywords were generated. He also had never used the advanced search form to search for Lin, and the department had no practice for checking search capability upon uploading new articles. Indeed, he admitted that the search capability may not have worked properly in 1999 or afterward. Additionally, he had no knowledge if any member of the public had actually accessed Lin from the library website.
Author’s web page. In its reply, petitioner also argued that Lin was publicly accessible because it was listed on a web page maintained by one of the co-authors of the Lin report. Petitioner submitted evidence showing the web page (with an affidavit from the Internet Archive) and a declaration from a library science expert who opined that the web page included an active link to the UCSD library website containing Lin. In addition, petitioner’s technical expert opined that a researcher would have sought online resources by going to web pages of other researchers (such as the co-author’s page).
Discussion: The Board ruled that the evidence failed to establish that Lin was publicly accessible. With respect to the UCSD technical reports library, the Board faulted petitioner for failing to provide evidence showing that a reasonably diligent searcher would have been able to locate Lin on the UCSD web site. As for the author’s web page, the Board found that the arguments and evidence were outside the permissible scope of reply, and in any event did not establish public accessibility. In support of these conclusions, the Board addressed several issues relating to the “touchstone” of public accessibility (i.e., whether the reference is disseminated or locatable), as discussed next.
“Technical accessibility” alone insufficient. According to the Board, the UCSD declaration established that the Lin report was uploaded to the library’s website as of November 23, 1999, because it described the university’s normal practice for receiving articles and uploading them to the library website. However, such “technical accessibility” was, by itself, insufficient to establish public accessibility. This result follows from Blue Calypso, which held that a technical report was not publicly accessible even though it was available to be downloaded over the Internet. Rather, the Board required evidence that Lin was disseminated or otherwise made available to the interested public – which required additional analysis.
No actual dissemination/access. As explained above, a reference retrievable over the Internet is not per se accessible. Thus the PTAB looked for actual dissemination of Lin to the interested public, but the record did not show any evidence that Lin was actually disseminated. Similarly, there was no showing that any member of the public actually accessed Lin once it was posted on the UCSD library website. Accordingly, there was no evidence of actual dissemination to the public.
Is it Locatable? The Board then turned to the question of whether a POSA, exercising ordinary diligence, could have found Lin on the UCSD library website, an issue which involved consideration of indexing, search capability, and other knowledge about the author or the report.
Insufficient indexing. The Board considered two types of indexing – by commercial search engines, and by the UCSD library. There was no evidence in the record to suggest that any commercial search engines had indexed the UCSD library website. The UCSD library had a search page with indexing that permitted users to view a list of technical reports by author or by year, but not by subject matter. The Board ruled that the evidence of indexing by author was insufficient because the authors of the Lin report were not sufficiently well known to enable a POSA to locate that report, distinguishing circumstances where the author was well known in the relevant technical field. Similarly, the evidence of indexing by year could not establish public availability in the absence of any evidence regarding the number of reports on the library website for the year of the Lin report; the Board was unconvinced that a POSA could locate the article by skimming through possibly hundreds of titles, most having unrelated subject matter. While the absence of subject matter indexing was not fatal to the question (i.e., subject matter indexing is not necessary to establish public accessibility), it was an important factor to be considered.
Insufficient evidence of search capability. While the UCSD library website allegedly permitted users to perform a keyword search based on author, title, or abstract, the evidence of such a search capability was deemed deficient. The UCSD declarant did not know how (or even whether) the search function operated or how keywords were generated, he never used the search function to locate Lin, and there was no practice to cross-check search capability upon uploading new articles. And other evidence demonstrated that the library website search function in fact did not function properly.
Insufficient evidence of other knowledge or access means. Petitioner’s expert testified that in 1999 POSAs were aware of and regularly used online libraries for research in computer science. But the PTAB rejected this as a basis for public accessibility, because this was merely evidence of general practice and not specific to the UCSD computer science department website. The PTAB distinguished Voter Verified, where the Federal Circuit found that posting an article on a website well known to the community interested in the relevant subject matter was a key factor favoring public accessibility. Here, in contrast, there was no evidence that the UCSD website was well known to the community. Similarly, the PTAB rejected arguments based on author reputation because there was no evidence that the authors of the Lin report were well known in the relevant field.
Finally, the Board addressed petitioner’s reply argument reply that Lin was publicly accessible because it was listed on a web page maintained by one of the co-authors of the Lin report. The Board first commented that the evidence and arguments regarding this web page were outside the permissible scope of reply and, therefore, could properly be ignored. However, the Board went on to conclude that the evidence was not sufficient to establish public accessibility. In particular, here there was no evidence that Lin was actually accessed via the author’s web page, that a POSA could have located the web page using an Internet search engine, or that a POSA would otherwise have known about the author’s web page
Implications of Ruling: As shown by the Board’s analysis in this case, petitioners will continue to face demanding requirements to establish that a reference qualifies as a prior art publication. The mere fact that a reference may have been posted to a website, or linked from another website, does not establish that the reference was publicly accessible. Rather, evidence tending to show public accessibility – such as actual dissemination or access, indexing, searching, or author reputation, specific to the reference, author, or website at issue – may be crucial to meeting the required level of proof. Petitioners should be prepared to provide detailed evidence showing a prior art reference was publicly accessible before the priority date of a challenged patent, such as the types of evidence discussed (but not present) in the Activision Blizzard case. Meanwhile, patent owners should take advantage of opportunities to challenge the public accessibility of references, particularly where the evidence and arguments in a petition are conclusory, and evidence of actual dissemination, indexing or other indicia of accessibility may be lacking. In such cases, challenges can be effective in a preliminary response and may derail a case before trial is instituted, as lack of public accessibility for a reference may be dispositive.