USPTO Extends Deadline for AI Inventorship Comments as Some Criticize Pannu Factors

“C4IP noted that [Section 103] establishes that ‘patentability shall not be negated by the manner in which the invention was made,’ which should extend to inventors’ use of AI computer systems.”

USPTOOn June 6, the U.S. Patent and Trademark Office (USPTO) published a notice in the Federal Register announcing that the Office would be reopening the public comment period for the development of inventorship guidance surrounding inventions developed by artificial intelligence (AI) systems. Although the USPTO will continue to collect public comments until the middle of June, public comments received from patent industry stakeholders so far are largely urging the agency to adopt inventorship guidance that provides a broad pathway to patentability for AI-assisted inventions.

This February, the USPTO issued guidance on inventorship requirements for AI-assisted inventions about one year after the agency first issued a call for public comments on the subject. The guidance follows the reasoning of the U.S. Court of Appeals for the Federal Circuit in Thaler v. Vidal (2022), which affirmed the USPTO’s denial of patent rights for an AI-assisted invention because it failed human inventorship requirements under U.S. patent law. While the guidance did not categorically exclude inventorship for AI-assisted inventions, it discussed the application of the joint inventorship test from the Federal Circuit’s 1998 ruling in Pannu v. Iolab Corp. to determine whether the agency should grant a patent on an AI-assisted invention.

C4IP: ‘Significant Contribution’ Factor is Improper in AI-Assisted Invention Context

The Council for Innovation Promotion (C4IP) wrote to express concern over the uncertainty created by the agency’s guidance, noting that the agency is wrong to start from the premise that the use of an AI system is different than the use of any other tool. While the agency’s guidance seeks to establish inventorship under Pannu’s “significant contribution” factor, C4IP noted that 35 U.S.C. § 103 establishes that “patentability shall not be negated by the manner in which the invention was made,” which should extend to inventors’ use of AI computer systems.

Case law from the Federal Circuit and its predecessor court establishes that humans can claim inventorship when assessing the efficacy of outputs developed by other systems, C4IP contends. In Life Technologies v. Clontech Laboratories (2000), the Federal Circuit found no inequitable conduct stemming from the undisclosed use of published computer analysis predicting biological activities in an altered enzyme prior to patenting the result. Similarly, a 1974 ruling by the U.S. Court of Customs and Patent Appeals (CCPA) in Silvestri v. Grant upheld inventorship for an individual appreciating a specific chemical species resulting from a set of reactions. This case law cuts against inventorship examples included in the USPTO’s guidance, which exclude inventorship based on an inquiry into AI contributions that is expressly prohibited by Section 103, C4IP argued.

U.S. Chamber: More Examples of Successful Inventorship Using AI Would Be Helpful

The U.S. Chamber of Commerce wrote to express that it agreed with the fundamental positions that AI-assisted inventions are not categorically unpatentable and that human contribution is required for patentability. Granting patent rights only when a human inventor contributes to the development of an AI-assisted invention would best promote innovation in AI systems.

While the U.S. Chamber was generally comfortable with the USPTO’s use of the Pannu factors in the AI-assisted invention context, the business advocacy group suggested that more inventorship examples would be helpful. This is especially true in the context of the second Pannu factor, which requires that inventors make a contribution that is not insignificant when measured against the full invention. This factor will likely become more important as generative AI systems become more involved in the process of developing inventions. The U.S. Chamber also requested further examples of AI prompts that would constitute a significant contribution by a human inventor.

Stephen Y. Chow: Pannu Factors Should Be Replaced by Touchstone of Human Conception

Patent attorney Stephen Y. Chow, Of Counsel with Hsuanyeh Law Group, wrote to oppose the USPTO’s injection of the Pannu factors into the AI-assisted invention context. Such factors, developed to answer post-patenting challenges to inventorship having limited review of human conception, introduce additional uncertainty into the appraisal of patents that involve increasingly automated tools. Further, the USPTO’s guiding principles could dissuade efforts by natural persons who recognize and appreciate inventive advances without being a person of ordinary skill in the art.

Chow urged the USPTO to retract the Pannu test in favor of a more complete focus on human conception. Tracing a line of reasoning from U.S. Supreme Court rulings dating back to the 18th century, Chow notes that current formulations of human conception appearing in more recent Federal Circuit case decisions focuses on the operative moment when an inventor’s mind forms a definite and permanent idea of a complete invention.

CSIS-SCSP: Patent Rights for AI-Assisted Inventions Are a Matter of National Security

The Center for Strategic & International Studies (CSIS) and the Special Competitive Studies Project (SCSP) submitted a joint comment to the USPTO reflecting many points raised by both groups in a commentary published jointly in late February. Maintaining robust patent protections for AI-assisted inventions are important not only for economic prosperity but also national security, as both groups acknowledge the high-stakes technological competition between the United States and China in the AI sector.

While CSIS and SCSP lauded the USPTO’s efforts to clarify some misconceptions around AI-assisted inventions, their comment outlined a series of key points that the agency must address more properly. In particular, the two policy institutions note that the USPTO must clearly define what constitutes an AI system as opposed to other computing tools. These organizations also contend that the USPTO’s application of joint inventorship law to the AI-assisted invention context is troublesome because it compares the levels of contribution from AI systems and humans, making it more difficult for a human to qualify as an inventor simply for the type of computing tool chosen to develop an invention.

Image Source: Deposit Photos
Author: roxanabalint
Image ID: 125703412 



Warning & Disclaimer: The pages, articles and comments on 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

Join the Discussion

One comment so far. Add my comment.

  • [Avatar for Anon]
    June 11, 2024 07:45 am

    There be a simple flaw here: 103 cannot elevate humans who simply do not meet the legal definition of inventor to BE the inventor.

    We’ve covered this many times now, even drawing the parallel to the ugly period of time in US history of slave owners trying to claim “to be the inventor” because they owned the slaves (as chattel) that WERE the inventors.

Add Comment

Your email address will not be published. Required fields are marked *

Varsity Sponsors

IPWatchdog Events

Patent Portfolio Management Masters™ 2024
June 24 @ 1:00 pm - June 26 @ 2:00 pm EDT
Webinar – Sponsored by LexisNexis
August 22 @ 12:00 pm - 1:00 pm EDT
Women’s IP Forum
August 26 @ 11:00 am - August 27 @ 5:00 pm EDT
IP Solutions, Services & Platforms Expo
September 9 @ 1:00 pm - September 10 @ 2:00 pm EDT
Webinar – Sponsored by Anaqua
September 19 @ 12:00 pm - 1:00 pm EDT

From IPWatchdog