AI slop has a signature style. At a glance, it’s easy to clock the hallmarks of an AI-generated poster: Crowded blocks of cliché-ridden text. Cutesy characters straight from the uncanny valley. Inexplicably explosion-filled backgrounds.

But Canva might have the solution to graphic design’s AI crisis. The platform’s newest feature reimagines AI design from the ground up, and it’s causing a stir on social media.

Canva Brush is a vision-language model built for design. Rather than taking a user’s prompt and spitting out a fully formed graphic, Brush takes designs step by step the way a human would, thanks to training on real design processes.

Brush is also built to combat one of the most frustrating aspects of AI image-generation. Often, asking for a single edit has an unintended domino effect, messing with unrelated aspects of a graphic that the user didn’t want to change at all. By keeping track of its interconnected decisions, Brush can isolate design aspects to make small changes. Its outputs are also fully editable for users who want to add a truly human touch.

A promo video for Brush shows the tool being used to create mock-ups for an ice cream brand, then generate staged product mock-ups and apply the same branding to different flavors (including a selection of Halloween pun-based cartons just in time for spooky season). The video has amassed more than 2.2 million views on X.

Introducing Brush: Our first vision-language model for design. 🖌️ Brush sees your design’s layout, layers and vibe the way you do, and edits alongside you.

Source link