AI and Trademarks: Rethinking Likelihood of Confusion
How might artificial intelligence (AI) reshape the way trademark professionals evaluate "likelihood of confusion"? This is one of the most important—and rapidly evolving—questions in trademark law. INTA’s recently released study, Exploring the Use of Artificial Intelligence in the Likelihood of Confusion Analysis, is the first major attempt to provide an answer.It takes a close look at how AI technologies—from machine learning and natural language processing to image and audio recognition—could be applied to one of the most fundamental critical aspects of trademark law: determining whether consumers are likely to be confused between two competing marks.The likelihood of confusion analysis is undertaken in three significant phases of the trademark lifecycle. The first use is made by a brand owner looking to adopt and likely register a new mark. The second is in intellectual property (IP) offices where trademark registration decisions include this analysis. The third area where the likelihood of confusion analysis becomes critical is when a dispute arises.Likelihood of confusion is certainly a legal test, but that legal test is rooted in the core functions of a trademark, that is how brands build trust with their consumers and how the brand owner’s investment in the trademark can be protected. Turning to the report, what makes it particularly timely is the context? Firstly, IP offices around the world are facing increasing application volumes and growing complexity, from multilingual word marks to logos, sounds, and even motion-based trademarks. At the same time, AI is already transforming how we search, compare, and analyze data. Importantly, the study does not advocate for or against AI. Instead, it offers a practical, forward-looking framework, highlighting both the opportunities and the risks, from improved speed and pattern recognition to concerns around bias, transparency, and over-reliance on automation. This episode of Brand & New is hosted by David Gooder, former Commissioner for Trademarks at the United States Patent and Trademark Office (USPTO), who also contributed to this report. To explore and help explain the report’s findings, he is joined by two colleagues who played a central role in developing this report.Lori Schulman is Senior Director of Internet Policy at INTA. Ms. Schulman has long been at the forefront of INTA’s work on digital policy issues, including the intersection of emerging technologies and IP. She was part of the core INTA project team that worked closely with FTI Consulting to shape this report, contributing both policy insight and practical perspective on how AI is already affecting the IP ecosystem. Jose Londoño is Associate Senior Director for External Relations and Policy Development at INTA. Mr. Londoño also served on the report’s internal project team and brings extensive experience working with IP offices and policymakers around the world. Mr. Londoño’s background in public-sector IP administration—which includes serving as Deputy Superintendent for Industrial Property at Colombia’s IP office—adds an important global and operational perspective to how AI could be implemented in practice. Related ResourcesAbout David GooderAbout Jose LondoñoAbout Lori SchulmanExploring the Use of Artificial Intelligence in the Likelihood of Confusion Analysis (INTA, June 2026)IP Office of the Future: Digital Transformation Think Tank Report (INTA, July 2026)INTA Board Resolution of Artificial Intelligence and Intellectual Property Rights Foundational Principles (INTA, May 2025)Likelihood of confusion (USPTO)