by David Hricik, Mercer Law School
Dennis pointed out that the Director has in Ecto World, LLC v. RAI Strategic Holdings, Inc, IPR2024-01280_paper_13_20250519 made it very difficult to obtain institution if a petition is based on art that was listed on an IDS even if it was not actually applied by an examiner, unless the art was listed on a “voluminous” IDS. That is a step back from the more real-world approach that had included examining whether the reference had actually been applied.
Dennis points out that avoiding “voluminous” (whatever that means) IDSs should help insulate patents from institution. Also reducing length is a statement by the Office that signing an IDS certifies that the practitioner has reviewed every reference listed. Specifically, In its Guidance on Use of Artificial Intelligence-Based Tools, the Office the Office expressed a concern that, while AI can be used to populate IDSs, doing so posed “the danger of increasing the number and size of IDS submissions to the USPTO, which could burden the Office with large numbers of cumulative and irrelevant submissions.” 89 Fed. Reg. at 25615-16. As a result, it noted the duty of reasonable inquiry included “reviewing each piece of prior art listed on the form” to determine if it was compliant with 37 CFR 11.18(b). Id. at 25616. Thus, review is required and clearly irrelevant and marginally pertinent information should be removed to avoid filing a paper for an improper purpose. Id.
That approach, interestingly, is somewhat similar to what OED Director Moatz suggested maybe 20 years ago, which was heavily criticized by practitioners as unrealistic. I am not sure the Office’s statement about reviewing each reference has achieved as much publicity now as it did back then, so stay tuned.
It seems to me that it could be argued that references found by AI are different than those found by a human searcher or those found in a search report in a related application in another country, because of AI’s tendency (at least on occasion) to produce, clearly, erroneous results more so that a human. Thus, it could be argued that this creates a higher obligation to review the references found by AI somewhat more closely than when found by other means. Despite the comments about the applicable rules being general and not being AI-specific, the higher propensity of AI to produce erroneous results is the overriding concern of ” Guidance on Use of Artificial Intelligence-Based Tools.”
There was a time when IDSs needed to be accompanied by a statement of the relevance of the references. However, the PTO decided to institute a more lenient standard to ensure that the Examiner receives all the art. It seems to me that lenient requirements for IDSs are better. While in general, not including cumulative references and “marginally” relevant references seem reasonable, sometimes a parent and continuation disclose different relevant-things (e.g., even if that is not supposed to be the case) and it can be difficult to foresee what reference an Examiner might find relevant as a secondary reference, for example. It could be argued that it is better to leave the decision of whether a reference is relevant to the Examiner, as long as there is at least a small possibility that the reference may be used in a rejection.