NEW YORK - The International Trademark Association (INTA) has released its latest research study, Exploring the Use of Artificial Intelligence in the Likelihood of Confusion Analysis. As intellectual property (IP) offices and professionals face increasing application volumes and growing complexity, this Report provides timely and deep insights into how AI could enhance efficiency, consistency, and data-driven analysis, while preserving the critical role of human judgment, according to the official website of INTA.

INTA chose to study likelihood of confusion because that analysis is a routine task central to trademark law and, importantly, assessed through the eyes of the “average consumer.” It is therefore a representative example of how artificial intelligence (AI) could be applied to subjective, as well as objective, AI legal analysis. The Association undertook this research in collaboration with business consulting firm FTI Consulting.

“AI is reshaping how brands are created, searched, examined, and protected, and INTA has an important role to play in helping the global IP community navigate this shift responsibly,” said INTA CEO Etienne Sanz de Acedo. “We are putting this work forward because likelihood of confusion sits at the heart of trademark law and as AI becomes more entrenched in legal and commercial decision-making. This Report provides IP offices and our members with a practical, forward-looking resource grounded in both legal analysis and technical understanding.”

INTA Chief Policy Officer Heather Steinmeyer served on the internal project team that drafted the Report. Underscoring why this issue is especially timely, she said: “AI has the potential to become an important decision-support tool for trademark examination, but its responsible use will depend on transparency, accountability, high-quality data, and continued human expertise.”

The Report spotlights and provides deep analysis into several key issues at the intersection of AI and trademark law, including the following:

• AI is poised to reshape trademark examination at scale: The Report finds that, as global filings continue to rise, AI can help IP offices and professionals analyze growing volumes of trademark applications, compare large datasets, and uncover potential conflicts faster.
•  
• AI isn’t replacing human judgment: While AI can enhance speed, consistency, and data-driven insight, the Report underscores that likelihood of confusion analysis remains deeply contextual, requiring an examiner’s legal reasoning, cultural understanding, and discretion. This is demonstrated in the context of determining the “average consumer,” as noted above.
•  
• Multimodal AI could unlock more sophisticated likelihood of confusion analysis: The Report highlights how AI tools may support visual, phonetic, conceptual, audio, and even motion-based comparisons, offering new ways to evaluate traditional and non-traditional marks in an increasingly digital marketplace.
•  
• Trustworthy AI requires guardrails: Emphasizing the need for explainability, auditability, continuous validation, and human oversight, the Report spotlights risks in the use of AI in likelihood of confusion analysis, including biased training data, model drift, opaque outputs, and over-reliance on automation.

“Written without the intention of advocating for or against AI in this context, the Report emphasizes that AI should function as a supportive tool, not a replacement for examiners or legal professionals,” emphasized Ms. Steinmeyer.

INTA unveiled the Report during a virtual launch event earlier today. Close to 500 participants from 60+ jurisdictions participated in the event, including IP office and government officials, IP professionals, academics, and members of the press.

The Report is part of INTA’s ongoing effort to support its members and the global IP community in navigating emerging technologies and their impact on IP law and practice.