Figma just made a quiet move that feels loud to anyone who’s ever exported a file to Photoshop for a five-minute fix. The company has folded three new AI-powered image-editing tools — Erase object, Isolate object, and Expand image — directly into Figma Design and Figma Draw, and put them in a single contextual image-editing toolbar so you don’t have to open another app every time a background needs cleaning up or a banner needs to be repurposed.

On paper, the features look familiar because they mirror functions designers have been using in Photoshop for years: erase (remove something and fill the gap), isolate (cut a subject out so you can move or style it independently), and expand (extend the image to fit a new aspect ratio). But the product-level idea here isn’t to out-Photoshop Photoshop — it’s to shorten the drag-and-drop commute between apps. If you live in Figma all day, the company’s bet is that “good enough” edits inside the canvas will save enough time and friction to make switching tools the exception rather than the rule.

Erase and Isolate piggyback on Figma’s existing lasso selection: you roughly draw around the thing you want to remove or lift out, then pick the action. In Erase mode, generative AI fills in what should belong behind the deleted subject (think: clone-and-heal but without the manual brushwork). In Isolate mode, Figma separates the selected object onto its own editable layer so you can reposition, mask, or export it as an asset. The workflow is deliberately tactile and “sketchy” — you don’t need pixel-perfect masks to get useful results. That’s the practical win: spend 30 seconds to fix an awkward logo or to pull a product out of a busy photo, then get back to layout decisions.

Expand image is the one that will save a lot of social media headaches. Rather than stretch or awkwardly crop when you need a different aspect ratio, the tool uses generative fill to extend the background so the composition still reads. You can drag the blue handles to enlarge the canvas and the AI will synthesize pixels that blend with the scene, which makes converting a horizontal hero into a vertical story or a square post way less painful. If that sounds like Photoshop’s Generative Expand, that’s because it is — but again, the point is doing it without leaving Figma.

None of this means Adobe is out of the picture. Photoshop still owns advanced retouching, color workflows, high-fidelity masking, and the exhaustive toolset pro photo editors need. Figma’s image tools are aimed at everyday product design and marketing tasks: cleanups, quick composites, and layout-friendly crops that keep the team working inside the same file. For many teams, that’s enough — especially when deadlines and cross-team handoffs matter more than pixel-level perfection.

There are practical limits and costs to keep in mind. Figma’s new image tools live behind what the company calls a “Full Seat” — meaning they’re available on Professional, Organization, and Enterprise tiers with AI-enabled — and AI actions consume credits, so teams should watch how they use heavy operations at scale. The rollout is starting with Design and Draw, and Figma says it plans to extend the image tools across more of its product line next year. In short, it’s available to teams willing to pay for full seats and who are ready to budget AI credits into their workflows.

Strategically, the update is part convenience play, part platform defense. Over the past year, Figma’s moved deeper into generative features and integrations — including efforts to marry different AI models and editing nodes into a more modular workflow — so adding image-editing primitives feels like the expected next step. The result is a stickier, more capable canvas: the fewer times designers need to bounce between Figma and a raster editor, the more value Figma can keep within its collaborative environment. That’s good for workflows and for Figma’s business.

For designers and product teams, the practical takeaway is simple: try the tools for quick fixes and layout changes, but keep a tested Photoshop workflow for pixel-critical work. Use Erase for rogue subjects in assets, Isolate when you need flexible composition elements, and Expand when you’re resizing for multiple channels. And because AI operations draw on credits, set a team policy — for example, let junior designers use edits for social posts and reserve heavy retouches or experimental batch runs for dedicated sessions so you don’t burn through credits on accidental tries.

The update won’t solve every image problem — generative fills occasionally stutter on complex textures or tight reflections, and the “good enough” edit won’t replace a meticulous retouch — but it changes the calculus. For routine image work that used to mean exporting, editing, and reimporting, Figma just removed several teeth from that process. Over time, as the company rolls these tools into more parts of its stack and refines the AI models behind them, expect fewer file-switching rituals and more of the design day lived entirely inside a single, collaborative canvas.

If you want to test things right away, open a copy of a real project and try erasing an element you’d normally send to Photoshop — watch how the fill behaves, and check the isolated asset for edge artifacts. Then do a quick expand to square or vertical and inspect the seams at full size; that will tell you when Figma’s in-canvas magic is a match for what you need, and when it’s still a convenience rather than a replacement.