USPTO to use AI tool for automated search pilot
The USPTO continues to invest heavily in AI capabilities; a new pilot program is designed to assess patent applications for potential “prior art” issues.
The U.S Patent and Trademark Office is turning to artificial intelligence to help automate the process of analyzing and sharing “prior art” issues with patent applications.
In an Oct. 8 notice in the Federal Register, USPTO announced plans to launch an “Automated Search Pilot Program.” The program will evaluate “the impact of sharing the results of an automated search prior to examination of an original, noncontinuing, nonprovisional utility patent application,” USPTO’s notice stated.
Sharing the automated search results to patent applicants “will provide the applicant with an earlier communication regarding potential prior art issues in their application.”
Prior art searches can help people understand what has already been invented and opportunities for innovation in advance of filing a patent application.
]]>
“The earlier communication may provide the applicant with an opportunity to assess prior art issues before substantive examination and make a more informed decision on how to proceed with the application,” USPTO wrote in the Federal Register. “The ASRN also provides a new pathway to identify relevant prior art for patent examiners to improve examination quality and efficiency.”
USPTO will accept petitions to participate in the pilot program starting Oct. 20. The agency expects to accept petitions as part of up to 1,600 patent applications.
The launch of the new pilot comes as USPTO continues to invest heavily in AI capabilities across a range of functions. The agency has been particularly focused on using AI tools to review documents and reduce administrative burdens on its staff.
Earlier this year, USPTO released a request for information seeking feedback on the potential for automated solutions, including AI, to improve the agency’s processes. The RFI specifically asked about the potential for AI to streamline USPTO’s prior art searches.
The RFI stated that “the exponential growth of prior art and tremendous pace of technological innovation make it increasingly more difficult to quickly discover the most relevant prior art.”
The automated search relies on an internal AI tool developed by the USPTO. The tool uses “contextual information to find similar information in publicly available documents located in a number of databases available to the USPTO,” the agency noted in the Federal Register this week.
The AI models that support the tool were trained on publicly available data.
]]>
“The training data excludes applicant, inventor, and assignee information because this information may introduce potential biases in the model,” according to the Federal Register notice. “The USPTO has implemented measures for the AI tool to ensure data security and maintain patent application confidentiality.”
USPTO’s goal with the pilot is to “evaluate the scalability of generating and mailing the ASRN, and acquire information to inform next steps,” the Federal Register noted.
Copyright
© 2025 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.