Tag: US Patent and Trademark Office

For All Patent/Trademark Practitioners: USPTO Provides Guidance for Use of AI in Preparing USPTO Submissions

The USPTO expounds a clear message for patent and trademark attorneys, patent agents, and inventors: use of artificial intelligence (AI), including generative AI, in patent and trademark activities and filings before the USPTO entails risks to be mitigated, and you must disclose use of AI in creation of an invention or practice before the USPTO if the use of AI is material to patentability.

The USPTO’s new guidance issued on April 11, 2024 is a counterpart to its guidance issued on February 13, 2024, which addresses AI-assisted invention creation process. In the new guidance issued on April 11, 2024, USPTO officials communicate the risks of using AI in preparing USPTO submissions, including patent applications, affidavits, petitions, office action responses, information disclosure statements, Patent Trial and Appeal Board (PTAB) submissions, and trademark / Trademark Trial and Appeal Board (TTAB) submissions. The common theme between the February 13 and April 11 guidance is the duty to disclose to the USPTO all information known to be material to patentability.

Building on the USPTO’s existing rules and policies, the USPTO’s April 11 guidance discusses the following:

(A) The duty of candor and good faith – each individual associated with a proceeding at the USPTO owes the duty to disclose the USPTO all information known to be material to patentability, including on the use of AI by inventors, parties, and practitioners.

(B) Signature requirement and corresponding certifications – using AI to draft documents without verifying information risks “critical misstatements and omissions”. Any submission for the USPTO in which AI helped prepare must be carefully reviewed by practitioners, who are ultimately responsible, to ensure that they are true and submitted for a proper purpose.

(C) Confidentiality of information – sensitive and confidential client information risks being compromised if shared to third-party AI systems, some of which may be located outside of the United States.

(D) Foreign filing licenses and export regulations – a foreign filing license from the USPTO does not authorize the exporting of subject matter abroad for the preparation of patent applications to be filed in the United States. Practitioners must ensure data is not improperly exported when using AI.

(E) USPTO electronic systems’ policies – Practitioners using AI must be mindful of the terms and conditions for the USPTO’s electronic system, which prohibit the unauthorized access, actions, use, modifications, or disclosure of the data contained in the USPTO system in transit to/from the system.

(F) The USPTO Rules of Professional Conduct – when using the AI tools, practitioners must ensure that they are not violating the duties owed to clients. For example, practitioners must have the requisite legal, scientific, and technical knowledge to reasonably represent the client, without inappropriate reliance on AI. Practitioners also have duty to reasonably consult with the client, including about the use of AI in accomplishing the client’s objectives.

The USPTO’s April 11 guidance overall shares principles with the ethics guidelines that multiple state bars have issued related to generative AI use in practice of law, and addresses them in the patent- and trademark-specific context. Importantly, in addition to ethics considerations, the USPTO guidance reminds us that knowing or willful withholding of information about AI use under (A), overlooking AI’s misstatements leading to false certification under (B), or AI-mediated improper or unauthorized exporting of data or unauthorized access to data under (D) and (E) may lead to criminal or civil liability under federal law or penalties or sanctions by the USPTO.

On the positive side, the USPTO guidance describes the possible favorable aspects of AI “to expand access to our innovation ecosystem and lower costs for parties and practitioners…. The USPTO continues to be actively involved in the development of domestic and international measures to address AI considerations at the intersection of innovation, creativity, and intellectual property.” We expect more USPTO AI guidance to be forthcoming, so please do watch for continued updates in this area.

Copyright © 2024 Womble Bond Dickinson (US) LLP All Rights Reserved.
For more news on the Use of AI in Preparing Patent Applications, visit the NLR Intellectual Property section.

PTO to Patent Examiners: Make Interpretation of Means-Plus-Function Claims Clear in the Record

On March 18, 2024, the US Patent & Trademark Office (PTO) issued a memorandum to patent examiners addressing means-plus-function and step-plus-function claim limitations and how to clearly articulate, in the prosecution record, the PTO’s interpretation of such claim limitations. The goal of the memorandum is to ensure consistency in connection with the examination of such limitations, provide both the applicant and the public with notice regarding the claim interpretation used by the patent examiner, and provide the applicant an opportunity to advance a different claim interpretation early in the prosecution.

As stated in 35 U.S.C. §112(f), “[a]n element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.” The memorandum does not suggest any changes in interpretation of the statute.

One aspect of the memorandum is to remind examiners of the resources and guidance available when examining means-plus-function and step-plus-function claim limitations, specifically MPEP §§ 2181-2187 and refresh training. In accordance with the guidance, the primary steps when examining such claim elements include:

  • Determining whether a claim limitation invokes § 112(f)
  • Ensuring the record is clear with respect to invoking § 112(f)
  • Evaluating the description necessary to support a § 112(f) claim limitation under §§ 112(a) and (b).

To determine whether a claim limitation invokes §112(f), the guidance instructs examiners to employ the three-prong analysis set forth in MPEP § 2181, subsection I. Using this analysis, recitation of the terms “means” or “step” in association with functional language, rather than structure, material or acts for performing that function, should be interpreted as claim limitations invoking § 112(f). However, where these terms are accompanied by structure, materials or acts for performing the function, § 112(f) is not invoked. On the other hand, a limitation reciting functional language along with a generic placeholder term instead of “means,” which fails to recite sufficiently definite structure for performing the function, would nonetheless invoke § 112(f), according to a proper analysis. Examples of such generic placeholders include “mechanism for,” “module for,” “device for,” “unit for,” “component for,” “element for,” “member for,” “apparatus for,” “machine for” and “system for.”

An important caveat in the memorandum states that “[e]stablishing the interpretation of § 112(f) limitations in writing during prosecution is critical in supporting the agency goal of establishing a clear prosecution record.” The guidance advises examiners that form paragraphs are available in support of meeting this objective, which serve to inform “the applicant, the public, and the courts . . . as to the claim construction the examiner used during prosecution. This further informs the applicant, the public, and the courts (and the PTO for any post-grant review procedures) as to how the examiner searched and applied prior art based on the examiner’s interpretation of the claim.”

The memorandum further emphasizes the need to evaluate whether claims under §112(f) meet the written description and enablement requirements of § 112(a) and the definiteness requirement of § 112(b). Regarding the latter, the specification must clearly disclose a structure that is clearly linked to or associated with the function, which would be understood by one skilled in the art to perform the entire recited function. Further, “[f]or computer-implemented § 112(f) claim limitations, the specification must disclose an algorithm for performing the claimed specific computer function . . . [and] sufficiency of the disclosure of the algorithm must be determined in light of the level of ordinary skill in the art.”

The memorandum further states that an indefinite § 112(f) claim limitation “based on failure of the specification to disclose corresponding structure that performs the entire claimed function will also lack adequate written description and may not be sufficiently enabled to support the full scope of the claim under § 112(a).” Thus, in any § 112(f) analysis, an examiner must determine whether the specification establishes possession of the claimed invention and whether sufficient information is provided to enable one skilled in the pertinent art to make and use the claimed invention.

For further details, see the memorandum here and the Federal Register notice here.

© 2024 McDermott Will & Emery
For more on Means-Plus-Function, visit the NLR Intellectual Property section.