Figma Bets on Design-Specific AI Agent as Quality Lags Speed

Figma identifies a practical challenge in product development: AI tools can speed up initial drafts, but this often leads to an increased time and effort to achieve the high design quality needed for release. A Figma survey in South Korea, where 51% of developers used AI on every project in the past year, found that 27% of respondents who reported AI affecting their productivity also experienced longer product launch times.

The core problem, according to Figma, is that general-purpose AI tools lack an understanding of design context. This deficiency necessitates extensive revision cycles; 62% of surveyed respondents indicated they needed six to ten rounds of revisions before AI output could be applied to a live service. Rene Lo explained that products meant for release demand high standards in user experience and attention to detail, which general AI often fails to meet without substantial human intervention.

Figma's response is "Figma Agent," an AI tool designed to operate within design files. It aims to create or revise screens by drawing on a team's accumulated design information and selecting appropriate AI models, including those developed in-house. Chase Wilson, Figma's director of developer marketing, stated that Figma Agent embodies the company's design know-how, positioning it as a solution specifically tailored to streamline the process of building and shipping new products by embedding design-specific intelligence.