The United States Patent and Trademark Office (USPTO) has extended its AI Search Automation Pilot program (ASAP!) to June 1, 2026. Originally scheduled to stop accepting new applications on April 20, 2026, the Office has now extended the program duration and expanded its scope "to obtain more information to evaluate the effectiveness of the program." Simultaneously, the USPTO has waived the petition fee for the program.

By way of background, ASAP! is a pilot program under which original, non-continuing, non-provisional utility patent applications filed under 35 U.S.C. § 111(a) (including Track One applications) undergo an AI-driven automated prior art search prior to examination. The search results are delivered to the applicant in the form of an Automated Search Results Notice (ASRN), accompanied by search strings that can be used to retrieve cited U.S. patents and publications in the Patent Public Search tool. Applicants can input these search strings into the Patent Public Search tool to query the cited U.S. patents and published documents. The ASRN does not require a response from the applicant. However, applicants who receive the ASRN may choose to file a preliminary amendment, request a deferral of examination, or expressly abandon the patent application to obtain a refund of the search fee and excess claims fees. The examiner ultimately responsible for the application may, at their discretion, decide whether to adopt the ASRN when conducting an independent search.

There are three aspects of this extension worthy of significant attention. First, the USPTO has increased the target volume of the program from approximately 1,600 applications (200 per Technology Center) to at least 3,200 applications (400 per Technology Center). Second, the Office will continue to accept applications until June 1, 2026, or the date on which each Technology Center has registered at least 400 applications accepted into the program, whichever is earlier. Third, and perhaps most impactful for cost-sensitive applicants: for all applications filed on or after March 23, 2026, the $450 petition fee is waived. As always, applications must be submitted electronically via Patent Center on the patent filing date, and application documents must be in DOCX format.

The USPTO's move to extend the program and waive the petition fee is intriguing. It suggests that, to date, the adoption of the program across various Technology Centers has been uneven, or that the adoption rate has been insufficient to support any evaluation of the program's effectiveness by the USPTO. In fact, as of the writing of this article, all Technology Centers combined have received only 169 applications, of which only 76 have been approved.

Although the fee is not high in absolute terms, it did constitute a material barrier for applicants who perceived the return on investment as questionable. Eliminating this fee effectively acknowledges that the initial cost-benefit balance was misaligned. Furthermore, applicants typically need to review the relevance of the references retrieved by ASAP!, which itself incurs administrative expenses such as time costs or attorney fees. Therefore, even with the waiver of the official fee, the program is not truly "free."

Overall, the fee waiver and the expansion of the acceptance scale will make applicants who were previously on the sidelines more willing to participate in the program. The advantage lies in providing a free automated search service that may (or may not) discover relevant prior art previously unknown to the applicant. For applicants filing in technical fields where the USPTO AI search tools possess a deep and well-classified corpus (such as software and electrical technologies), the ASRN may indeed be very useful.

There is another development worth noting. At the University of San Diego Patent Law Conference held recently, it was reported that USPTO Director Squires announced that a second-generation AI search tool will be launched in the near future. According to Squires, the core improvement of the new version lies in the fact that it will conduct searches directly against the claims of the filed application. The version currently used in the pilot conducts searches based only on the patent specification. This situation is unexpected yet significant.

Any prior art search should cover and focus on the claims, as the claims define what is actually used to examine novelty and non-obviousness. Searches based solely on the specification often only detect prior art related to the overall disclosed embodiments rather than technology that matches the specific claim limitations asserted by the applicant. Consequently, the practical value of the current version of the program is itself questionable. However, if the second-generation tool is implemented as scheduled, the quality of the ASRN will be significantly improved.