USPTO Seeks Potential AI-Powered Solutions for Improving Prior Art Searches

The U.S. Patent and Trademark Office is conducting market research to identify vendors that can support its plan to deploy artificial intelligence to improve systems for prior art searches.

In a request for information posted on SAM .gov, the USPTO stated that prior art searches play a crucial role in patent examination processes to issue reliable patent rights; however, locating relevant prior art on current search systems is becoming increasingly difficult due to the rapid growth of prior art and fast pace of technological advancements, FedScoop reported.

The RFI consists of two phases. The first phase requires interested companies to submit answers to questions regarding their business experiences and capabilities related to mitigating bias in AI models and deploying and maintaining machine learning models with patent data sets. Under Phase 2, selected companies will be given the opportunity to demonstrate their solutions in a virtual event anticipated to be held between Sept. 29 and Oct. 13.

Responses to the RFI are due Sept. 11. Questions regarding the notice should be submitted by Friday.

The RFI said the USPTO has already employed AI-based patent searches but it plans to use more advanced technology for enhanced patent search capabilities and systems