Make any image editable with Canva’s Magic Layers

What if an image was not just an image, but an editable template? This is the principle of Canva’s new Magic Layers technology, which uses AI to interpret an image and transform it into an editable file.

Canva’s Magic Layers uses the company’s AI design model to analyze an image. The technology promises to separate parts of the image into editable layers, for example separating the bride and groom from a wedding photo. If there was a caption or overlay (“Congratulations, Mike and Diane!”), Magic Layers technology promises that the text can also be separated and edited.

If you create or edit a file in Microsoft Paint or Adobe Photoshop, all layers end up merged into a single final image that is exported as a file. Canva’s Magic Layers takes the opposite route, trying to make sense of what needs to be separated and edited.

This is a big turning point from an AI perspective, with implications for everything from design to misinformation. It’s also unclear how effective the technology will be at distinguishing individual elements of the scene and separating them into layers.

Canva says the Magic Layers technology should work with just about any JPEG or PNG file you can upload to its software, at least in beta. More capabilities will be added in the future. The image file is then uploaded to the Canva editor, where you can edit it.

Magic Layers begins rolling out in public beta today in the US, UK, Canada and Australia, with global availability to follow, Canva said.