Figma releases AutoFlow 2.0 with real-time AI-driven interaction mapping

Design · 5 min read

AutoFlow 2.0 uses a background inference engine that analyzes frames, layers, and component names to propose likely user flows and state transitions. Designers can accept, tweak, or reject generated arrows and conditions; the tool also suggests microcopy for CTA states and validation messages. The generator operates incrementally, updating flow suggestions as designers modify the canvas.

A key addition is a simulation mode that runs a lightweight behavior model to surface dead-ends and common navigation loops. This lets teams catch UX issues earlier by seeing how pseudo-users might click through prototypes. Figma says simulation logs and annotations can be exported to product analytics tools to align design intent with data.

AutoFlow 2.0 also introduces a plugin API for third-party tools to tap into the flow graph. Early adopters include testing platforms that convert generated flows into test cases. The update reflects a broader push to embed generative assistance directly into the authoring surface, not just offer image or copy generation.