Rules and Key Cases for Patent Practitioners Working on AI Patent Applications |  Sheppard Mullin Richter & Hampton LLP

Rules and Key Cases for Patent Practitioners Working on AI Patent Applications | Sheppard Mullin Richter & Hampton LLP

On September 22, 2022, the United States Patent and Trademark Office (USPTO) directed patent practitioners to current case law and sections of the Manual of Patent Examination Procedures (MPEP) as reminders while patent practitioners continue to work in the artificial intelligence (AI) technology space. A summary of these recalls (and links to more information) are provided here.

MPEP sections you should know – especially for AI inventions

Two main areas of concern for applications that relate to AI inventions are (1) patent subject matter admissibility and (2) enabling disclosure. Highlighted sections included MPEP 2106, MPEP 2181and MPEP 2173.05(g).

MPEP 2106 provides general guidance on subject matter eligibility, including the definition that the claimed invention must belong to one of four legal categories, the claimed invention must also be considered patentable subject matter (e.g., claim should not be directed to a judicial exception unless the allegation as a whole includes additional limitations in an amount significantly greater than the exception). The section also provides a flowchart explaining how examiners should analyze claims to determine if they relate to patent-eligible subject matter (under Step 1, Step 2A Stream 1, Step 2A Stream 2 and Step 2B) , and also provides a handful of helpful examples of eligible and ineligible claims. For example, specific to AI, MPEP 2106.03 describes that products that do not have a physical or tangible form, such as information (often called “data by itself”) or a computer program by itself (often called “software in itself”). se”) when claimed as a product without any structural narrative, are not directed to any of the legal categories and therefore do not mention the subject matter eligible for the patent.

MPEP 2181 provides general guidance for considering limitations of means plus function (35 USC 112(f)) to avoid a claim that is indefinite. In these cases, the means plus function analysis may be invoked by using the terms “means” or “step” in a claim, or “generic placeholder” terms are used instead of “means” or “step”. “. To overcome this construction, the applicant may (1) present sufficient evidence to establish that the limitation of the claim sets forth sufficient structure to perform the function claimed to avoid construction under Section 112(f); or (2) alter the claim limitation in a way that avoids interpretation under § 112(f) (eg, by reciting sufficient structure to perform the function claimed).

MPEP 2173.05(g) discusses functional limitations that do not invoke 35 USC 112(f), but may still render the claims indefinite. In this section, the claim states a characteristic “by what it does rather than by what it is” (for example, as evidenced by its specific structure or specific ingredients). Unlike the means-plus-function claim language which applies only to purely functional limitations, the functional claim often involves the recitation of a structure followed by its function. For example, the claim may cite a tapered spout (the structure) that “allows[ed] several popcorn kernels to cross at the same time” (the function). The applicant should be careful not to draft indefinite claims to state functional limitations.

Key PTAB and USPTO Decisions on AI Petitions

As of the date of the meeting, two decisions illustrate the current AI patent landscape. The decisions are: Ex parte Hannun (formerly Ex parte Linden), 2018-003323 (1 April 2019)that applies the Patent Eligible Guidance (PEG) 2019 to an invention “improving speech-to-text transcription”, and In re Appl. No. 16/524 350 (“DABUS”), according to which the quality of inventor must be limited to natural persons, and not to AI/machines. Additional details on each case are provided below.

In Ex parte Hannun, the patent in suit claimed a system and method for improving speech-to-text transcription. The PTO Examiner rejected the claims for allegedly not directed to patentable subject matter. For example, the PTO Examiner asserted that the claims were merely directed to a “mathematical relationship/formula” and also to “certain methods of organizing human activity”, while asserting that “since man can listen to an audio file and transcribe the audio data into textual data which can all be done mentally.Applicant appealed, and the Board agreed, that the claim was to patentable subject matter for several reasons. For example, in As part of the Stage 2A Stream 1 analysis, the Board asserted that “[w]Although the transcription can generally be performed by a human being, the claims here are directed to a specific implementation comprising the steps of normalizing an input file, generating a set of jitter audio files…” The Commission also asserted that “…claims do not include fundamental economic principles or practices, business or legal interactions, management of personal behavior or relationships or interactions between people…” so as not to cover the abstract idea of ​​“some ways of organizing human activity. Additionally, the Board asserted that “the claims use predicted character probabilities to decide on a transcription of the input audio, which the examiner, relying on the specification, determines using a mathematical formula . Namely, the Examiner identifies that the specification discloses an algorithm for obtaining the predicted character probabilities… The algorithm or mathematical formula, however, is not recited in the claims. Although the analysis could have stopped at the first step 2A, the Board also considered that “the claims of the present application include specific features which have been specifically designed to achieve an improved technological result” and “provide improvements to this technical area” (under Step 2A, Stream Two). The Board has also considered the PTO reviewer’s rejection at Step 2B. For example, the PTO Examiner concluded that the claims included “no additional element which amounts to much more than a judicial exception”, but failed to provide sufficient factual evidence, thus remedying another ineffectiveness of the rejection.

In In re Appl. No. 16/524 350 (“DABUS”), the plaintiff attempted to claim a machine as the inventor of a patent application. For example, the Application Data Sheet (ADS) listed a single inventor “DABUS” as the first name and “(Artificial Intelligence Generated Invention)” as the last name. The transferee was listed as “Stephen L. Thaler” (legal representative of DABUS). The “inventor’s statement” stated that the invention had been conceived by a “creativity machine” named “DABUS” and that he was to be named as the inventor in the ‘350 application. The USPTO issued a first notice of filing of the missing parts of the non-provisional application which stated that the ADS “does not identify each inventor by legal name” and an additional $80 for late submission of the oath or the declaration of the inventor. The plaintiff filed a petition under 37 CFR 1.181 requesting a review of the notice and rescission of the notice as unwarranted and/or void. The USPTO then issued a second notice of filing of the missing portions of the non-provisional application and denied the applicant’s motion. The plaintiff filed a request for reconsideration of the decision to deny the plaintiff’s motion. The USTPO held that the “inventor” must be a natural person and cannot encompass machines (e.g., neural networks) by reviewing case law, USC sections, and various USPTO rules cited in the MPEP, all of which refer to natural persons and pronouns.

Training examiners on subject eligibility for AI

The USPTO has published various examiner training materials for examining AI inventions. Examiner training materials are available here. These training materials cover in particular current guidance on patent subject matter eligibility (e.g. PEG guidance before 2019, PEG guidance after 2019, nature-based and life science products, and a selection of court decisions).

Particular attention is paid to PEG 2019 Example 39, which analyzes a patent claim relating to a “Method of training a neural network for facial detection” and confirms that the claim states a patent-eligible subject matter. For example, the preamble of the claim stated a “computer-implemented method for training a neural network for facial detection.” The claim items included:

  • collecting a set of digital facial images from a database;
  • applying one or more transformations to each digital facial image, including mirroring, rotation, smoothing, or contrast reduction to create a modified set of digital facial images;
  • creating a first training set comprising the collected set of digital facial images, the modified set of digital facial images and a set of digital non-facial images;
  • train the neural network in a first step using the first training set;
  • creating a second training set for a second training stage comprising the first training set and non-facial digital images that are incorrectly detected as facial images after the first training stage; and
  • training the neural network in a second step using the second training set.

Under Step 2A Prong One, the USPTO asserted that the claim does not cite any of the judicial exceptions listed in the 2019 PEG, so the claim is directed to patent-eligible subject matter. In particular, PEG 2019 asserts that the claim does not mention any mathematical relationships, formulas or calculations. While some of the limitations may be based on mathematical concepts, the mathematical concepts are not set forth in the claims. Also, the claim does not mention a mental process because the steps are hardly performed in the human mind. Finally, the claim does not state any method of organizing human activity such as a fundamental economic concept or the management of interactions between people. Thus, the claim is admissible because it does not mention a legal exception and the analysis does not go to stage 2A Part 2 or to stage 2B.

Similar Posts

Leave a Reply

Your email address will not be published.