Figma’s 2026 AI report: AI is changing how teams work together, not just how fast

Figma released its third annual AI report around Config 2026 in San Francisco in late June 2026. It draws on three years of data, 8,403 survey responses and 639 qualitative interviews with designers, developers and product managers.

This year it covered 10 markets: the US, UK, Japan, Germany, France, Canada, Australia, Brazil, India and South Korea, with the last three added for the first time. There is no data from Thailand or other ASEAN markets, so it should be read as an industry trend rather than a picture of the Thai market.

Figma’s headline conclusion is that the biggest change is not speed or output but how teams work together.

Some 41% of respondents say AI meaningfully changes how their team collaborates, up from just 7% two years ago. Role crossover shows it most clearly. Designers doing development work rose to 41%, from 21% in 2025. Developers doing design work rose to 60%, from 44%.

Where people work is shifting too. 76% do at least half their work on the canvas, and about six in ten spend most of their time there. One in five developers would rather start a project on the canvas than in code. According to iTWire, 77% of product managers also work on the canvas at least half the time.

Figma groups company approaches to AI into four patterns: unified (36%), directive or top-down (27%), grassroots (20%) and nascent (18%).

A common worry is that AI will make design less valuable. The survey points the other way. 90% say design is at least as important as before AI, and nearly six in ten say it is more important. iTWire reports that respondents at the most advanced AI-adopting companies are 25% more likely to say design has grown in importance.

The reasoning holds up. When anyone can produce screens or code quickly, the scarce skill becomes deciding what should be built, what is easy to use and what fits the brand. That is design work.

There is also the question of distinctiveness. Work that AI produces quickly tends to look alike, because it learns from common patterns. Brands that want to stand out still need people who frame the brief, choose between options and refine the result. This is our analysis, not a figure from the report.

Set against the 2025 report as summarised by TechRadar, the picture has moved. That survey of 2,500 people found 78% said AI made workflows more efficient, but fewer than half said it made them better at their job, and 69% of designers were satisfied with AI tools compared with 83% of developers. This year the focus has shifted from “is it faster?” to “how is teamwork different?”

At Config 2026, Figma announced tools that follow this trend: Code layers for working with code directly on the canvas, cloning repositories and syncing changes back; Motion for animation; WebGPU-powered Shaders; and agent skills that let teams turn repeated work into shared skills. Availability varies by feature and is rolling out over time.

In our view, this data answers a question Thai business owners often ask: “If AI can build a website, do we still need a designer?” Industry insiders say AI speeds up building, but decisions about user experience and brand consistency carry more weight, not less.

For in-house teams, more crossover between designers and developers makes a clear design system even more important. Without agreed colours, type and components, AI-assisted work from several people will not fit together.

There is also the question of how to adopt AI. Many Thai organisations let staff try tools on their own. That starts quickly, but without shared guidelines at some point, results become scattered.

There is no need to change every tool at once, but it is worth laying foundations so that AI-assisted work stays consistent.

If you hire an agency, ask where the team uses AI and who makes the user experience and brand decisions. A clear answer is a sign the team uses AI to speed up work, not to replace thinking.

• Build or review a design system with components and rules both people and AI can use
• Have designers and developers work in the same files from the start of a project
• Decide which work AI can draft and which needs a designer’s judgement
• Trial new features on small projects and compare time and quality with the old way