The U.S. Patent and Trademark Office (USPTO) has announced the rollout of DesignVision, an advanced AI‑powered image search tool now integrated into the examination system for U.S. design patent applications. According to the Official Gazette notice published July 16, 2025, DesignVision enables examiners to perform federated searches of U.S. and foreign industrial design and trademark databases—spanning over 80 registers—using only image inputs. Results are ranked by visual similarity, allowing faster and more precise prior art discovery. The notice clarifies that this tool is intended to augment existing textual search methods, not replace them.
This move reflects the USPTO’s commitment to streamlining design patent examination, reducing examination delays, and increasing consistency across search operations. Examiners will still rely on established text‑based tools and non‑patent literature, but will now have AI‑driven image search as an enhanced capability. Unique features of the tool include the ability to select up to seven uploaded images as a search query, user-controlled weighting of visual features, user-controlled focus on specific features of the searched image, and text and classification filters. Applicants will be notified when DesignVision has been used to examine their application, and the confidentiality of the searched designs will be maintained within the USPTO’s system.
The public notice emphasizes that DesignVision follows a broader agency effort to modernize IP processes through responsible AI integration. Notably, DesignVision is not unique to the U.S. Many other international patent and trademark offices have adopted similar AI image‑based search tools, recognizing their power to accelerate examination timelines and enhance examination quality. The USPTO’s public notice and implementation thus underscore a growing trend: smart image recognition tools are becoming a standard component of design patent work worldwide.
