Trademark classification goes agentic with USPTO's announcement of "Class ACT" assistant

The United States Patent and Trademark Office announced the release of an artificial intelligence agent that handles pre-processing tasks for trademark applications. The Trademark Classification Agentic Codification Tool, or Class ACT, assigns international classes to unclassified applications. It also assigns design search codes and pseudo marks that make the records searchable. This capability will enhance searches by examining attorneys and practitioners. Director John A. Squires said classification and design search coding that takes five months could now take five minutes or even five seconds. The office is providing specific, task-directed AI agents to tackle the toughest and most information-intensive aspects of pre-examination. Acting Chief AI and Data Officer Rob Hayes said the Trademark Office is setting the standard with faster results, higher quality, and happier stakeholders. Hayes joined the office as a senior advisor from a senior post at X. Director Squires congratulated Hayes and Acting Trademark Commissioner Dan Vavonese for addressing the delays that slow applications from the outset. Trademark applications with logos, designs, unconventional spelling, or no international class have historically made searching difficult. The surge in applications caused the manual addition of codes and classifications to take several months. AI now provides this information immediately at a high level of accuracy, though humans at the USPTO still review it. Senior Legal Advisor Kathleen Cooney-Porter said the benefits are available to examining attorneys and the public almost instantly. Acting Trademark Commissioner Dan Vavonese said employees can focus on applying their experienced judgment and reason to the substantive issues in examination. More AI trademark solutions are coming soon.