The USPTO published a request for comments (RFC) on April 30th, focusing on how advancements in artificial intelligence (AI) may impact the USPTO’s assessment of patentability governing (i) what may qualify as prior art and (ii) determining the level of skill of a person having ordinary skill in the art (PHOSITA).[1] Accordingly, the USPTO is seeking written comments from the public on or before July 29th, to help evaluate and develop guidance as well as provide technical advice to Congress in considering future legislation.[2] This RFC expands on President Biden’s 2023 executive order (see https://www.bakerlaw.com/insights/white-house-instructs-uspto-to-provide-guidance-on-ai/) and the USPTO’s Inventorship Guidance for AI-Assisted Inventions published February 13th (see https://www.bakerlaw.com/insights/uspto-delivers-inventorship-guidance-on-ai-assisted-inventions/).[3],[4]
Section II of the RFC explains the prior art framework of 35 U.S.C. 102 and poses questions about whether AI-generated disclosures may qualify as prior art.[5] For a disclosure to qualify as a printed publication, it must have been publicly available so a PHOSITA exercising reasonable diligence could locate it as of the effective filing date of the application.[6] Because vast numbers of AI-generated disclosures may be autonomously created without any human involvement, questions arise whether a PHOSITA may have been able to locate it.[7] Further, because a presumption of operability exists for prior art disclosures, questions arise whether AI-generated prior art disclosures potentially devoid of human review should be given the same presumption of operability and enablement in view of AI’s ability to create false or misleading subject matter, e.g., hallucinations.[8]
Section III of the RFC describes the current legal framework of 35 U.S.C. 103 involving a determination of obviousness from the perspective of a PHOSITA.[9] Section III also explains obviousness determinations that require resolving the level of ordinary skill in the art.[10] The Federal Circuit has specified factors affecting the level of skill determination, including known problems encountered in the prior art, prior art solutions, how fast innovations are made, and the sophistication of the technology involved.[11] Further, Section III discusses written description requirements under 35 U.S.C. 112, inherency determinations, and claim construction viewed in terms of the level of skill in the art and, therefore, how they may be potentially implicated by the utilization of AI.[12]
Section IV of the RFC lays out specific questions soliciting public commentary based on the issues described in Sections II and III. Specifically, the questions under subsections IV.A. and IV.B. address the impact of AI on prior art and the impact of AI on determining patentability involving a PHOSITA, respectively.[13] Further, subsection IV.C. poses a series of questions directed to updating the Inventorship Guidance or legislative changes.[14]
Section IV.A: AI’s Effect on Prior Art[15]
- Does the Patent Act presume or require that prior art has been authored or published by a human?
- What AI-generated disclosures would be pertinent to patentability determinations? How should such disclosures be made available to the public?
- Is there a duty to notify the USPTO if a party submits a disclosure or evidence they know was generated by AI?
- Should AI-generated disclosures be treated different than other disclosures, such as (a) in considering the extent of human contribution to the AI-generated disclosure, (b) considering the extent of the presence of incorrect information, and (c) considering the operability, enablement and public accessibility of the disclosure?
- Does AI-generated prior art create an undue barrier to the patentability of inventions?
Section IV.B: AI’s Effect on the Level of Skill in the Art[16]
- Is a PHOSITA a natural person? How does AI affect the level of skill of a PHOSITA as AI becomes more prevalent?
- How should the USPTO determine which AI tools are commonly used in a particular art?
- How does the availability of AI to a PHOSITA impact how they would construe the claims or what is well known in the art?
- How is obviousness affected by AI in evaluating (a) analogous art; (b) a suitable rationale; (c) predictability in the art; and (d) commercial success, long-felt but unsolved needs, failure of others, simultaneous invention, unexpected results, copying, etc.?
- How does the recency of the information used to train an AI model impact the PHOSITA assessment when the assessment may be focused on an earlier time?
- How does the availability of AI to a PHOSITA impact enablement?
Section IV.C: USPTO Guidance and Legislative Considerations[17]
- What guidance on these questions from the USPTO would be helpful?
- In what other ways does the proliferation of AI impact patentability?
- Are there any laws or practices in other countries that address these questions?
- Should title 35 of the U.S. Code be amended, and why?
[1] See Request for Comments Regarding the Impact of the Proliferation of Artificial Intelligence on Prior Art, the Knowledge of a Person Having Ordinary Skill in the Art, and Determinations of Patentability Made in View of the Foregoing, 89 FR 34217 (Apr. 30, 2024) [hereinafter RFC].
[2] Id.
[3] Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence, Executive Order 14110, 88 FR 75191 (Nov. 1, 2023).
[4] Inventorship Guidance for AI-Assisted Inventions, 89 Fed. Reg. 10043-51 (Feb. 13, 2024).
[5] Id.
[6] RFC at 34218-19; see also MPEP 2128.I; In re Wyer, 655 F.2d 221, 226 (C.C.P.A. 1981).
[7] RFC at 34219.
[8] RFC at 34219; see also MPEP 2121.
[9] RFC at 34219; 35 U.S.C. 103.
[10] RFC at 34219; Graham v. John Deere Co., 383 U.S. 1, 17 (1966).
[11] RFC at 34219; In re GPAC Inc., 57 F.3d 1573, 1579 (Fed. Cir. 1995).
[12] RFC at 34219.
[13] Id. at 34219-20.
[14] Id. at 34220.
[15] Id. at 34219-20.
[16] Id. at 34220.
[17] Id.
[View source.]
