According to its Design Patent Dashboard, the United States Patent and Trademark Office (USPTO) currently employs 286 design patent examiners, has pendency from filing to first action of 16.7 months, a que of 76,014 unexamined design patent applications, and is advertising to hire additional design patent examiners.

Through a Request for Information (RFI) entitled Artificial Intelligence for Design Patent Applications (ACQ-22-1598) posted in November 2022, the USPTO is conducting market research to assess the maturity of Artificial Intelligence (AI) capabilities as they specifically relate to “image searching” during design patent application prosecution.

The RFI notes that “design patent prosecution” refers to the series of interactions between inventors, practitioners, and Design Patent Examiners that can result in a design patent. When the Applicant submits their ornamental design for an article of manufacture, the USPTO assigns a Design Patent Examiner to review the application disclosure and claim to determine whether to grant a design patent. One of the most important tasks of the Design Patent Examiner is to substantively review what is disclosed in the application and compare it to the already existing references called prior art.

The RFI states that prior art is defined as all information that has been made available to the public in any form before a given date that might be relevant to a design patent application's claim of novelty and non-obviousness.

Design Patent Examiners are challenged with searching all of an ever increasingly complex and vast corpus of human knowledge in a limited amount of time. They must read and understand the patent application, and perform an extensive prior art search, citing references to determine new contributions made or to explain the basis for rejecting the application.

Image Searching has become an increasingly challenging aspect of design patent examination because:

- Innovators, artists, and designers publish new designs every day. There is an exponential growth of information to be searched. While most internet searches are similar to a “seek and find” book, a prior art search for a disclosed design is more similar to looking for a needle that did not already exist in an ever-growing haystack.

- Appearance matters, and text searching does not cast the best net in catching the most relevant references. The appearance of a design is its own best description. While image search databases scour the internet for similar images, the examiner still needs to be able to find unique visual features within larger designs that are not readily available.

- Design Patent Examiners require a secure database and data exchange in order to upload sensitive, unpublished, proprietary data in order to retrieve results.

- Finally, because the Design Patent Examiner’s time is limited, they need better results, not more results. The components of a better priors art search result include the United States Patent Classification (USPC) design specific designation or CPC designation for non-design US patents or publications, ranking or relevancy of the results, combined with relevant visual features to help Design Patent Examiners make their determination of novelty and non-obviousness.

In Guiding Designs, we reported on updates to the USPTO Manual of Patent Examining Procedure (M.P.E.P.) to cite case law which draws additional significance to the identification of the article of manufacture in the title and claim. It will be interesting how image-based searching could reveal a landscape of similar appearing designs, but if the results are not limited to the specific article of manufacture identified in the language of the title and claim, or for the same intended purpose, or from the same field of endeavor, will that really be useful for examination?

Currently design patent applications are classified into 33 subject matter classes, with subclasses enumerating more specific type and shape characteristics – see the USPTO’s Classification of Design Patents. According to current practice, design patents are classified as in the first Design class that contains subject matter most pertinent to the subject matter claimed and in the first pertinent subclass or indented subclass that describes the specific function, intended use, or ornamental features of the design claimed.

The USPTO is not alone in exploring AI technologies to assist IP examination; the World Intellectual Property Organization (WIPO) maintains an Index of AI Initiatives at IP Offices. Notably, Australia, China, EUIPO, Japan, Singapore, WIPO itself, as well as others, are developing image-based AI classification and search tools to support design examination,

George Raynal, Principal

george.raynal@designlawgroup.com

© 2023 Saidman DesignLaw Group, LLC