The United States Patent and Trademark Office (USPTO) published its report entitled Public Views on Artificial Intelligence and Intellectual Property Policy (the Report) addressing the impact of artificial intelligence (AI) technology across the intellectual property (IP) landscape. The public submissions included comments from a broad range of experts, including foreign patent offices, IP practitioners, academia, and companies. This two-part article focuses on the views expressed in the Report with respect to the impact of AI on trademark law and practice. Part 1 focuses on the use of AI in the practice of trademark law, while Part 2 (in the next issue of the OBA's Newsletter) addresses if and how the current American trademark legal landscape will be impacted by the use of AI in the marketplace.
When assessing how AI will impact the practice of US trademark law, the common view in the Report is that the use of AI in trademark searching will have a positive impact on the industry. One area of improvement is utilizing AI software to assist IP registries to conduct trademark examinations more efficiently. A well-trained AI tool can streamline the examination process by identifying all perceptible design elements in a proposed mark to produce results that are more comprehensive, consistent, and accurate than the current process at the USPTO, where the examining attorney manually assigns numerical design codes to conduct a search and determine the risk of confusion with registered marks.
On the other side of the spectrum, concerns are raised in the Report that the use of AI in trademark examinations will have a negative impact on applicants. Trademark applicants may end up facing more rejections on the basis of the likelihood of confusion compared to the traditional design code method because AI-generated search results are likely to find similar marks that would have otherwise been unidentified during examination, thereby making the process for successful registration more costly and difficult. A balance should therefore be struck between the benefit of a more efficient and effective examination process for the USPTO and the risk of a potentially more burdensome registration process for the applicants when evaluating the use of AI.