Canva’s new editing tool adds layers to AI-generated designs

Canva has introduced a new feature that separates flat image files and AI-generated visuals into multi-layered, fully editable designs. the Magic classes The tool launches in public beta today in the US, UK, Canada and Australia, allowing design components such as objects, text boxes and other graphics to be individually selected while preserving the original layout.

“Following advances from our AI research team, we’re introducing Magic Layers so anyone can take a flat photo and turn it into a fully editable design within Canva,” said Cameron Adams, Canva’s chief product officer. he said in the press release. “There’s no need to start over, or to figure out the right vector. Generation is just the beginning – true creative freedom comes from the ability to edit without losing momentum.”

The feature isn’t just designed to work with AI-generated works — Magic Layers currently supports any single-page PNG or JPEG file, with “expanded development capabilities.” However, the focus on how it can be used in AI images is not surprising, given how aggressively Canva has been pushing its generative AI tools over the past couple of years. The goal here is to prevent Canva users from having to re-render an AI-generated image if only a small portion needs to be modified.

It’s also a step above what other creative software providers offer. Generative AI tools in Adobe applications like Photoshop and Express will add generated elements to their own distinct layer that’s easy to separate from the main design, for example, but they don’t currently provide a way to automatically split the entire image into layers.

Magic Layers may give creators more manual control over how flat images are edited, at least compared to asking an AI assistant to make small adjustments, but they may also make it harder to distinguish AI-generated designs from those made from scratch. Since image generators are (currently) limited to producing flat designs, displaying layers of elements is one way artists can prove that they’ve made something without simply describing it to the machine.