GAO Says USPTO Falls Short on Oversight, Patent Quality Efforts

“Without stronger internal controls, clearer goals, and more accurate performance data, the agency risks perpetuating a system where patents are granted too quickly and challenged too often.”

GAOThe U.S. Government Accountability Office (GAO) has issued a sharply critical report concluding that the United States Patent and Trademark Office (USPTO) has failed to effectively assess and manage its initiatives aimed at improving patent quality. The April 2025 report—prepared in response to a request from the Senate Judiciary Committee’s Subcommittee on Intellectual Property—found that the agency’s efforts to improve patent examination are hampered by poor planning, inadequate oversight and misaligned performance metrics.

Despite modest reforms since the GAO’s 2016 reports, USPTO examiners say they are still incentivized to prioritize speed and quantity of output over examination quality. The report identifies multiple gaps in how the agency measures both examiner performance and overall patent quality and calls on the USPTO to adopt more transparent, outcome-based approaches.

Background and Context

The study was initiated in light of continuing concern that low-quality patents may be inhibiting innovation and inflating legal costs. According to the GAO, the USPTO granted roughly 365,000 patents in FY2024, out of over 527,000 applications. Yet roughly 40% of litigated patents are ultimately invalidated-raising questions about examination rigor.

The GAO has previously flagged many of these issues in reports GAO-16-490 and GAO-16-479. While the USPTO has implemented some recommendations since then, the 2025 report suggests the agency has not made enough progress on meaningful, system-wide reform.

The GAO’s conclusion is clear: if the USPTO is to restore stakeholder confidence in the reliability of granted patents, it must rethink how it defines, measures, and enforces examination quality. Without stronger internal controls, clearer goals, and more accurate performance data, the agency risks perpetuating a system where patents are granted too quickly and challenged too often.

Focus on Output Undermines Quality

The GAO conducted six focus groups with nearly 50 patent examiners representing most of the agency’s technology centers. Examiners overwhelmingly reported that their performance evaluations continue to be driven by production quotas-measured in “counts” for completing review tasks-rather than thorough prior art searches or legally sound evaluations of patentability

This output-driven culture, GAO found, contributes to examiners shortchanging time-intensive work such as evaluating complex prior art. And while the USPTO has made efforts to emphasize patent quality in examiner appraisals since FY2021, the report found no evidence that these changes improved examination standards.

Inconsistent Oversight and Weak Metrics

The GAO found that the USPTO lacks a formal structure for developing and evaluating pilot programs designed to improve patent quality. Of the 14 pilot programs reviewed, half did not include any outcome-based evaluation to determine scalability or long-term value. This, GAO said, reflects missed opportunities to improve future pilot efforts.

Even core supervisory quality reviews—the main tool for assessing examiner performance—suffer from reliability issues. Supervisors are not required to use random sampling in selecting cases for review, can exclude identified errors, and in some cases assign passing scores even when all reviewed work contains mistakes.

This approach, the GAO found, “likely overstates examiners’ true adherence to quality standards and may not encourage consistent high-quality examination.”

Missing the Forest for the Trees

While the USPTO tracks compliance with each statutory patentability requirement individually—such as novelty, nonobviousness and disclosure—it does not set or report a goal for how many granted patents meet all four requirements together.

For FY2023, while individual requirement compliance ranged from 92% to 98%, the share of patents that satisfied all four requirements at once was just 84%. The GAO concluded that this gap distorts perceptions of patent quality and urged the agency to establish and communicate an overall compliance goal.

The report also noted that the USPTO has not implemented any metrics to evaluate the scientific or economic value of issued patents, despite having a strategic goal to better measure innovation. The GAO pointed to widely recognized proxies such as citation frequency or real-world usage metrics as opportunities to improve accountability.

Eight Recommendations

The report included eight recommendations to the USPTO aimed at improving oversight and transparency in the agency’s quality efforts:

  1. Evaluate ongoing initiatives to determine whether they are actually improving examination quality.
  2. Formalize pilot program development using best practices for testing, evaluation, and scalability.
  3. Update guidance on supervisory reviews to ensure better sampling and eliminate review bias.
  4. Establish an overall patent quality goal that measures full statutory compliance.
  5. Communicate that goal publicly to provide a clearer picture to stakeholders and Congress.
  6. Assess and adopt indicators of patent value, including scientific and economic metrics.
  7. Ensure more accurate measurement of examiner performance through independent reviews.
  8. Improve internal controls and consistency monitoring to reduce variability in examination outcomes.

Looking Ahead

These findings call to mind concerns raised in a 2022 House IP Subcommittee hearing, where the GAO’s preliminary report revealed that a majority of PTAB judges surveyed felt pressured to conform to USPTO management’s preferences in America Invents Act trials. At the time, lawmakers and stakeholders warned that such internal influences could erode judicial independence and called for structural reforms. The latest report broadens the focus from adjudication to examination, underscoring that issues of oversight and performance metrics span the USPTO’s core functions.

While the USPTO has taken steps in recent years to modernize its tools and enhance examiner training, the GAO report makes clear that foundational issues remain unresolved. As Congress and the innovation community turn more attention to patent quality, the agency will face increased pressure to implement more rigorous and transparent standards.

This article was updated on 5-8 to clarify that reexaminations were not an area identified by examiners as a challenge.

Share

Warning & Disclaimer: The pages, articles and comments on IPWatchdog.com do not constitute legal advice, nor do they create any attorney-client relationship. The articles published express the personal opinion and views of the author as of the time of publication and should not be attributed to the author’s employer, clients or the sponsors of IPWatchdog.com.

Join the Discussion

12 comments so far.

  • [Avatar for Anon]
    Anon
    May 15, 2025 02:43 pm

    Thank you francis – I will say that your approach is refreshing in contrast to the likes of MaxDrei, who more typically barges in with an EPO Uber Alles approach that misses many of the chosen distinctions between our Sovereigns.

  • [Avatar for francis hagel]
    francis hagel
    May 15, 2025 11:14 am

    @Anon 12 May

    I agree. It is especially important that the examiner defines the scope of the initial search in light of the entire application. For example, it is common that the specification identifies a field of use as of special interest to the applicant but none of the initial claims recites it. Including the field of use as a claim limitation is a likely fallback position if the initial broad claims are anticipated by prior art outside the field of use. It is then clearly desirable for the initial search to encompass the field of use highlighted by the specification.

    On the broad topic of patent quality and avenues toward improvement, the practice of the EPO’s SQAPs (“Stakeholder Quality Assurance panels”) deserves attention notwithstanding any anti-Europe sentiment on your side of the pond. As explained in the Annex pages 78-84 of the EPO 2023 Quality Report (https://link.epo.org/web/general/annual-review-2023/en-quality-report-2023.pdf ), a number of EPO cases selected at random is audited for quality issues by SQAP panels each made up of a mix of 2 EPO senior examiners and 2 European patent attorneys (private practice or corporate) appointed by users associations. Each panel is specialized in a certain field and different subjects are reviewed. The striking result is that in most cases, there is consensus between EPO experts and external auditors as to the quality issues and suggestions for improvement. Also, quality issues identified by the panels are specific and relevant to the EPO practice as well as to practitioners’ concerns. I have signalled the EPO 2023 SQAP report in epi-information 2024-3 (https://information.patentepi.org/issue-3-2024/epo-report-sqaps-2023-findings.html).

  • [Avatar for Anon]
    Anon
    May 12, 2025 04:39 pm

    Well stated, Mr. Lewis.

    Kudos as well, Miss Burke.

    Miss Trujillo, not sure that I can agree that any ‘front loading’ has the effect that you indicate – especially if examination of the entire application (rather than, say, a claim element plug and chug) were done. Perhaps I would re-phrase the fault as to a combination of lack of diligent front-loading, and application of the phrase, “If you cannot find the time to do it right in the first instance, when will you find the time?”

  • [Avatar for David Lewis]
    David Lewis
    May 12, 2025 02:32 pm

    “Yet roughly 40% of litigated patents are ultimately invalidated-raising questions about examination rigor.”

    Really? I would hope that litigators, infringers, and patent holders are smart enough to choose their fights and only litigate patents that they think they can have a reasonable chance of invalidating. If so, the 40% number will not change no matter how high the quality of patents is. What might be more indicative of patent quality is what percentage of patents are litigated (no matter the outcome).

    “Supervisors … in some cases assign passing scores even when all reviewed work contains mistakes.” Is that really so terrible? Are all mistakes equal? Shouldn’t the number of mistakes be an issue. The point being, I don’t see a problem with an examiner getting a passing score, even if all work reviewed has errors, depending on the number of mistakes and nature of the mistakes.

    “The report also noted that the USPTO has not implemented any metrics to evaluate the scientific or economic value of issued patents, despite having a strategic goal to better measure innovation. ” Should the PTO really be in the business of evaluating the scientific or economic value of a patent? There is nothing in the statues that require patent to have economic of scientific value – nor should there be. Isn’t the idea to encourage the filing of patents to encourage the sharing of ideas. If some ideas are going to be derided as not economically or scientifically significant, inventors will just keep their ideas to themselves (even those that are economically or scientifically significant, because no one really knows which ideas are significant).

    Perhaps consistency how Office Actions between Examiners, Art Units and Groups should be looked at?

    It is curious that the PTO is being asked to decrease its work force, when it sets its fees to cover its expenses.

    As many above have commented, don’t expect higher quality if you are not going to give examiners more time and higher more examiners so you can give examiners more time without increasing the pendency.

    Presumably, the reason why the PTO continues to emphasize quantity over quality is because Congress and/or the Administration focus on and reward productivity, whether or not it is an express performance metric by which the US PTO is measured.

  • [Avatar for Patrick Kilbride]
    Patrick Kilbride
    May 9, 2025 10:50 am

    This report reflects the fatally flawed understanding that has taken hold in American law and politics of IP as a mere stimulus to innovation and creativity. When a patent claim is evaluated based on a subjective assessment of its “inventiveness,” patent “quality” will always be questioned. The only consideration that should matter is whether the claim reflects fundamentally new work. Whether that work has inventiveness, utility, or value should be purely for markets to decide – others will rely on the new work, or they won’t. The singular focus of patent examination reform should be providing the technological tools that enable IP administrators to understand the validity of a claim to new work and where that work fits within the existing body of knowledge, technology and prior art.

  • [Avatar for Doreen Trujillo]
    Doreen Trujillo
    May 9, 2025 10:18 am

    Footnote 22 of the report is telling.

    “22 Because the USPTO wants its examiners to issue complete and thorough first actions, examiners get approximately three times the number of counts for the first action as they do for subsequent office actions”

    And that is the major problem in my view.

    The USPTO is front-loading examination credit. That process presumes applicants are filing claims that need little to no amendments to be allowed. The only claims that meet such a presumption, however, are very narrow ones. Such claims are easy to design around to avoid literal infringement. And a determination of infringement under the doctrine of equivalents is all but nonexistent.

    According to the report, the Examiners think that applicants lead with broad claims because they are trying to game the system. I do not believe that to be the case in the majority of instances. Applicants lead with the broadest claims they hope to obtain in view of the art they are aware of and their understanding of the statute requirements, and knowing that a finding of equivalents is not likely.

  • [Avatar for James Packard Love]
    James Packard Love
    May 9, 2025 07:46 am

    Complaints about patent quality are as old as the patent office. For example, the New York Times , on May 5, 1860: “Were the Patent Office a mere registry of inventions, granting monopolies for all things till they were upset by the Courts, the result would be precisely the same.”

  • [Avatar for Brian]
    Brian
    May 8, 2025 08:50 pm

    Wants better quality but isn’t willing to pay for it, typical. A reoccurring theme these days. Want better quality, give examiners more time to do their jobs. Want the pendency to also go down, hire more examiners. Want better metrics for supervisors, give them the time to do it. Can’t have it both ways without paying for it. What I expect will happen is the avg examiner will now have more roadblocks, not be given more time, then will still be expected to churn out enough cases to move the backlog down. Attorneys always laugh when they find out how many hours we get per office action. They think it’s a really insane as they get close to twice more billables.

  • [Avatar for Julie Burke, PhD]
    Julie Burke, PhD
    May 8, 2025 06:55 pm

    “And while the USPTO has made efforts to emphasize patent quality in examiner appraisals since FY2021, the report found no evidence that these changes improved examination standards.”

    To be clear, in 2021, the USPTO *told* Congress that they revised the patent examiners’ performance and appraisal plan (PAP) to create a roadmap to quality.

    Let’s take a close look at changes made to the quality element of the 2021 PAP.

    In 2010 and 2015, the quality element weighed 35% of the overall rating.
    After 2021, the quality element was reduced to 30% of the overall rating.

    In 2010 and 2015, examiners could be rated Outstanding if they made up to 4.49% errors.
    In 2021, the number of errors permitted for an Outstanding rating was increased to 6.49% errors.

    In 2010 and 2015, the number of Quality Major Activities assessed for error in a primary examiner’s rejection or allowance was 16.
    In 2021, the number of Quality Major Activities assessed for error was reduced to 12.

    In 2015, the ADVANCED standard was used to evaluate whether a primary examiner properly determined compliance with 35 USC 102, 103 or 112.
    In 2021, the USPTO evaluated patent examiners’ determinations under 102, 103 or 112 under a BASIC standard.

    In 2010 and 2015, examiners must comply with Indicia 1-3 in SUBSTANTIALLY ALL actions in order to obtain a Commendable rating.
    In 2021, examiners must only comply with Indicia 1-3 in a MAJORITY OF their actions to received a Commendable rating.

    No further changes were made to the quality element in the 2024 PAP.

    The only change made in the 2024PAP, as compared to the 2021PAP was to give patent examiners up to one hour other time for reviewed IDSs which are more than ten pages long.

    Does these changes look like “emphasizing patent quality”?

    Why would anyone expect these changes to improve patent quality?

    For more, please see my IPWatchdog article from 2021 titled “Does the USPTO’s Roadmap to Improved Patent Quality Lead to Lake Wobegon?”

    https://ipwatchdog.com/2021/07/01/usptos-roadmap-improved-patent-quality-lead-lake-wobegon/id=135152/

  • [Avatar for J]
    J
    May 8, 2025 11:52 am

    Using litigation invalidated patents in such a dumb metric. Many applications are abandoned. Many patents are licensed. Litigated patents are edge cases where 2 outside parties disagree, but also a case where both parties are spending 100x what the USPTO gets to spend on an application.

  • [Avatar for None]
    None
    May 8, 2025 11:32 am

    Do you want higher quality patents or lower pendency? If you want both, start calling your representatives and tell them to end the hiring freeze and let off the chokehold on training/oversight hours.

  • [Avatar for Truth]
    Truth
    May 8, 2025 08:47 am

    Quality = 1/Production

    Instructions to PTO:

    Increase production to cut backlog, but also increase *quality. Lose up to 15% of the workforce.

Varsity Sponsors

Industry Events

Certified Patent Valuation Analyst Training
May 28 @ 9:00 am - May 29 @ 5:00 pm EDT
2026 WIPO-U.S. Summer School on Intellectual Property
June 1 @ 9:00 am - June 12 @ 1:45 pm EDT

From IPWatchdog