Figma is embedding an AI assistant directly inside its collaborative design canvas, turning the design workspace into a place where human teams and AI agents work side by side. Instead of switching to a separate prompt tool, designers can ask the AI to create or modify work directly within the same environment where projects are built and reviewed.
The new capability represents Figma’s most direct integration of generative AI into its core workflow: a design agent that operates on the canvas itself, understands the context of the design file, and performs actions based on simple natural‑language prompts.
Users interact with the assistant by describing what they want in plain language. From there, the agent can perform a range of design tasks without manual layout work.
Typical actions include:
For example, a user might prompt the assistant to create a landing page layout, tweak a button style, or generate several layout alternatives to explore design directions faster. The goal is to reduce repetitive production tasks so teams can focus more on ideas and decision‑making.
One of the more unusual features is the ability to run several agents at the same time.
Instead of issuing a single request and waiting, teams can launch multiple AI agents to work on different tasks simultaneously—for instance generating layout options, adjusting components, or testing variations in parallel. This approach effectively treats AI as multiple collaborators operating within the same shared canvas.
The system isn’t just a generic chatbot. Figma describes it as a purpose‑built design agent trained to understand interface elements, layout structures, and design systems.
Because it operates inside the canvas itself, the assistant can interpret existing frames, components, and styles in a file. That contextual awareness allows it to remix designs, automate routine tasks, and follow existing design system rules rather than producing unrelated visuals.
Figma has also been integrating AI coding environments from partners such as Anthropic and OpenAI, enabling tools like Claude Code and Codex to interact with the design platform. These integrations help connect design artifacts with code‑generation workflows.
The AI agent is rolling out initially within Figma Design, the company’s main interface design product. The rollout is gradual and begins with a beta release, with broader availability planned later.
During the beta period, the feature does not consume AI usage credits, though credits are expected to apply when it reaches general availability.
Embedding AI agents in the canvas is part of a broader push by Figma to bring design and development workflows closer together.
The company has been expanding its platform beyond traditional UI design tools. At its Config conference, Figma introduced products such as:
These additions are designed to help teams move from concept to production within a single platform.
Figma’s move also comes amid rising competition in AI‑driven design tools.
Companies such as Canva and Adobe are building their own generative design capabilities, while new AI products can create layouts, marketing graphics, or prototypes directly from prompts. The emergence of conversational design tools and prompt‑based prototyping has increased pressure on traditional design platforms to integrate AI deeply into their workflows.
By embedding an agent directly in the collaborative canvas, Figma is betting that AI should enhance the existing team‑based design process rather than replace it.
The bigger idea behind Figma’s assistant is that AI becomes another participant in the design workspace.
Instead of replacing designers, the agent helps explore ideas faster, generate alternatives, and handle repetitive production work. Designers still make the decisions, but they can now delegate experimentation and iteration to AI agents operating inside the same canvas.
If the approach works, the design process could start to look less like manual layout work and more like directing a team of creative agents that generate, test, and refine visual ideas in real time.
Studio Global AI
Use this topic as a starting point for a fresh source-backed answer, then compare citations before you share it.
Figma’s new AI assistant works directly inside its collaborative canvas, letting users generate or edit designs with natural‑language prompts, automate variations, and even run multiple AI agents simultaneously while...
Figma’s new AI assistant works directly inside its collaborative canvas, letting users generate or edit designs with natural‑language prompts, automate variations, and even run multiple AI agents simultaneously while... The assistant is powered by design‑focused models and builds on integrations with AI tools like Claude Code and Codex, part of Figma’s broader push to connect design workflows with code.
The feature is launching first in Figma Design in beta and reflects growing competition among AI design platforms from companies like Canva and Adobe.
Loading comments...
Comments
0 comments