USPTO Launches “ASAP!” Pilot: AI-Powered Prior Art Search Enters Patent Examination

What ASAP! Means for Your Patent Practice

The U.S. Patent and Trademark Office (USPTO) has launched an artificial intelligence-driven pilot program that fundamentally changes how early-stage prior art discovery works. The “Artificial Intelligence Search Automated Pilot Program” (ASAP!) is now accepting applications through April 20, 2026, and offers patent practitioners a new strategic tool for prosecution planning. Rather than waiting months to discover what prior art an examiner will cite, ASAP! delivers an automated prior art report within weeks of filing, giving applicants a rare window to shape examination outcomes before they formally begin.

Participation is voluntary, and the automated search results notice (ASRN) is not a formal office action—it carries no response obligation. But for practitioners, ASRN information represents actionable intelligence: early visibility into the prior art landscape before the examiner’s official action, an opportunity to file preliminary amendments while still in the pre-examination window, or even a decision point to strategically abandon or defer unpromising applications without incurring examination fees.

How the AI Works: Architecture and Search Scope

ASAP!’s AI system derives its power from a combination of classification data and contextual analysis. The system extracts information from the Cooperative Patent Classification (CPC) system, then analyzes the application’s specification, claims, and abstract to identify the technical domain with precision. This contextual framing—not just keyword matching—allows the AI to pull the most technically relevant references rather than generic hits.

The search scope is expansive: U.S. Patents, U.S. Pre-Grant Publications, and the Foreign Image and Text database (FIT), which aggregates image and text materials from foreign patent offices. Results are ranked by relevance and capped at ten documents per application. A critical design feature for fairness: the AI was trained on publicly available patent data, with applicant, inventor, and assignee information deliberately excluded. This exclusion is intended to reduce algorithmic bias and ensure that the prior art discovery is driven by technical content, not by who filed the application.

Eligibility and Program Capacity

Not all applications qualify. ASAP! accepts only original, noncontinuing, nonprovisional utility applications filed under 35 U.S.C. § 111(a). Continuing applications, plant patents, design patents, reissue applications, and international 371 applications are ineligible. Notably, Track One applications do qualify. The application must be filed in DOCX format and must be accompanied by form PTO/SB/470. Applicants must enroll in the Patent Center electronic Office Action program before submission.

The program has fixed capacity: a minimum of 1,600 applications total, with at least 200 slots reserved per technology center. This limitation means that not every eligible application will be accepted; ASAP! operates on a first-come, first-served basis within its filing window (October 20, 2025 through April 20, 2026). Practitioners should verify current enrollment status by contacting AutomatedSearchPilot@uspto.gov or calling 571-272-7704 before finalizing filing strategy.

The Petition Fee and Documentation Requirements

Participation requires payment of a petition fee under 37 CFR § 1.17(f). While the fee amount is modest compared to the value of early prior art intelligence, it remains a material cost in portfolio management. The required form, PTO/SB/470, must accompany the application. All filings must be in DOCX format—PDF or paper submissions will not be accepted. These technical requirements are non-negotiable and should be verified before submission to avoid eligibility rejection.

Strategic Opportunities: What You Can Do with an ASRN

Receipt of an ASRN opens three distinct strategic paths. The first is proactive amendment: armed with concrete prior art references, you can file preliminary amendments while prosecution still awaits examiner assignment. This approach allows applicants to narrow claims, clarify scope, or refocus dependent claims based on known references. Unlike responding to an office action, preliminary amendments avoid formal rejection records and can streamline prosecution.

The second option is deferral. ASRN information may reveal that the invention faces significant prior art obstacles that warrant deeper analysis, client consultation, or design-around investigation. Applicants can request examination deferral to buy time for these deliberations without losing filing priority. This is particularly valuable for complex technologies where a few months of additional development or prior art analysis can shift the patentability outcome.

The third option is strategic abandonment. If ASRN references suggest the application faces insurmountable prior art rejections, applicants can opt for express abandonment and recover the application fee. For marginal applications in lower-value technologies, this option can free resources for stronger candidates elsewhere in the portfolio. The ability to make this decision based on actual prior art—rather than speculation—represents genuine cost control.

What ASRN References Do and Do Not Do

An important limitation: documents in the ASRN are not automatically cited by examiners and do not appear on the patent’s face unless the examiner cites them during examination or the applicant submits them through an Information Disclosure Statement (IDS). The ASRN is an intelligence tool, not a binding citation. This distinction matters for prosecution strategy. If you file amendments based on an ASRN reference, you may choose not to formally disclose that reference to the examiner, preserving the possibility of unanticipated office actions. Conversely, if references seem particularly relevant to patentability or claim scope, filing an IDS after ASRN receipt can ensure they appear in the examination record.

The Broader Implications: AI in Patent Examination

ASAP! represents the first formal integration of AI-driven prior art search into the USPTO examination pipeline. This is significant because historically, prior art discovery has been bottlenecked by examiner workload and keyword-dependent search strategies. AI systems, particularly those trained on high-dimensional patent data, can identify technical relationships that keyword searches miss. In complex fields like software, biotechnology, and mechanical engineering, this capability could materially improve the quality and completeness of prior art retrieval.

For practitioners, the most immediate benefit is predictability. A substantial source of examination surprise—unexpected prior art references—becomes less likely when applicants receive early notice of the examiner’s likely references. This shifts patent prosecution from reactive firefighting to proactive strategy. Preliminary amendments, deliberate claim narrowing, or strategic withdrawal become informed decisions rather than reactive responses. The downstream effect is faster prosecution, fewer office actions, and more controlled expenditure on examination.

However, AI limitations deserve scrutiny. These systems excel at pattern recognition within well-documented technical domains but may struggle with genuinely novel inventions that fall outside established classification paradigms. An invention that combines techniques from disparate fields, or that challenges conventional problem-solving approaches, might not be fully captured by an AI system trained to recognize known technical relationships. Patent practitioners should not assume ASRN results represent the universe of relevant prior art; they remain one source of intelligence among many.

Program Evaluation and Long-Term Outlook

The USPTO plans to evaluate ASAP! using the Government Accountability Office’s “Leading Practices for Effective Pilot Design,” a rigorous framework for pilot assessment. This means the program will be formally tested on key metrics: search precision and recall, applicant satisfaction, impact on examination timing, potential algorithmic bias, and effects on claim allowance rates. The results will heavily influence whether and how the USPTO expands AI-assisted examination across the agency.

Practitioners should view ASAP! as both an immediate tactical opportunity and a signal about the USPTO’s longer-term direction. If the pilot succeeds, AI-powered examination features will likely expand beyond prior art search to other examination functions. If significant limitations emerge—such as poor recall in certain technologies, uneven results across technology centers, or bias concerns—the program may be narrowed or abandoned. Either way, participation data and outcomes will inform patent office policy for years to come.

The enrollment window closes on April 20, 2026. Given program capacity constraints and the strategic value of early ASRN reports, practitioners should assess eligible applications now and prioritize enrollment for those most likely to benefit from pre-examination prior art intelligence. For questions or enrollment verification, contact the ASAP! team at AutomatedSearchPilot@uspto.gov or 571-272-7704.