Insights

Thought Leadership

Patently Enabled October 2025 – Smarter Searches, Tougher Standards: USPTO Launches DesignVision

The U.S. Patent and Trademark Office (USPTO) recently launched DesignVision, an artificial intelligence (AI)-powered image search tool developed to improve the efficiency of design patent examiners and address bottlenecks at the USPTO. While promising to improve design patent examiner performance, DesignVision is likely to increase rejection rates and complicate prosecution for design patent applicants, as more comprehensive prior art searches yield more novelty rejections. Applicants might therefore expect to pay higher professional fees for design patent prosecution as a requisite cost for obtaining stronger design patents.

AI can scour far and wide

DesignVision works by comparing images submitted in design patent applications to extant images in more than 80 design patent databases, including those of the USPTO, the World Intellectual Property Organization and the European Union Intellectual Property Office. Upon upload to DesignVision, newly submitted images are compared to database images, and DesignVision can even rank database images by degree of similarity to the input image based on examiner-selected visual characteristics. Such visual pattern recognition now significantly exceeds human capabilities and is a forte of AI. DesignVision should therefore allow examiners to identify more relevant prior art, resulting in more art-based rejections.

Can AI be trusted?

To promote transparency and ensure accountability, the USPTO currently tracks DesignVision use. When an examiner implements DesignVision, the design applicant is notified of the selected image query, and the USPTO creates a public record that an AI tool has been utilized. Notably, DesignVision is a design-directed component of the USPTO's Patent End to End (PE2E), a fully integrated web-based suite of both AI and non-AI tools developed to assist examiners during the application review process.

USPTO's earlier AI-powered intellectual property tool: Similarity Search

The USPTO's introduction of DesignVision follows past rollouts of other AI-powered patent tools, most notably Similarity Search – a 2022-implemented language-based, AI-powered search tool similar to DesignVision; however, rather than utilizing images as the query, Similarity Search uses natural language to search U.S. and foreign patent databases for similar language in prior art patent applications. As with DesignVision, patent examiners using Similarity Search can adjust parameters to prioritize certain language or concepts, thereby enhancing the search process. An examiner's use of Similarity Search creates a public record in an application's search history. The USPTO's PE2E also includes Similarity Search, which is now a required component of the utility patent application review process.

AI at the USPTO is the future

AI search tools have recently outpaced traditional USPTO examiner search capabilities; AI tools are so efficient at churning through huge amounts of information that they have become critical tools for USPTO examiners. With DesignVision, design patent applicants can now expect more rigorous design application examination. Even if initial iterations of DesignVision require improvement, the USPTO appears committed to widespread AI adoption, and effective upgrades should be expected. In short, AI in the design patent realm is here to stay – in much the same way as it inhabits the utility patent space – and DesignVision will likely produce both welcome and unwanted consequences for design patent applicants.