Adobe’s experimental tool can identify an artist’s work online or on a tote bag

Metavives October 16, 2024

Project Know How builds on Adobe’s work with Content Credentials. | Illustration by Alex Castro / The Verge

One of Adobe’s most notable experiments this year could help combat misinformation and ensure artists are credited for their work, no matter where it appears online or offline. Announced during the Sneaks segment at Adobe Max, Project “Know How’ is an in-development tool that can link ownership of an image or video across any online platform, and a multitude of real-world surfaces like posters, tumblers, and textiles.

Project Know How builds on Adobe’s Content Credentials tech, which applies a digital tag to track where a piece of content has been posted, who owns it, and if/how it’s been manipulated. Providing an image or video has Content Credentials applied, the tool can help creators establish ownership over their content even if that authentication metadata has been stripped out. The demo I saw, while early in development, managed to display the Content Credentials data on an image just by recording it on a camera, even on a texture-heavy object like a tote bag.

Here’s an example of Project Know How in action — a live camera feed on the left scans an object, and then displays a still “snapshot” (on the right) once the tool identifies that the image has Content Credentials attached, which can be clicked on to be directed to the information page.

This could solve a multitude of gripes that artists face around how to protect their work. It’s a common issue for art and designs to be posted to online platforms like Pinterest and X without crediting the original creator, and therefore limiting the benefits of such exposure. While the creative community can address this manually by leaving comments or community notes that credit the original artist, Know How provides another level of protection that’s difficult for bad actors to bypass.

The tool also provides other benefits — if a video has been clipped or taken out of context, Know How can fetch the original full-length video in the state it was first published in. Adobe says that local governments have, or are looking to, adopt Content Credentials, so this could offer additional protection against manipulated videos being used to push political misinformation online.

*Project Know How was able to find the original video from the shortened clip on the left, and reveal that one kid had actually been edited out.*

Adobe’s sneaks aren’t guaranteed to be released as fully realized products, but with the aggressive push the company has made into its Content Credentials initiative of late, I’d be surprised if some variant of these features aren’t adopted eventually. There’s no word on when that could happen, but demonstrating that it’s technically possible right now is an impressive feat that could help to reassure creatives concerned with proving ownership of their works.