Griffin Wooldridge: AI tools for UX designers — a complete 2026 workflow

Published on January 3, 2026, this video by Griffin Wooldridge walks through a complete AI-assisted UX design workflow. Wooldridge is a working UX designer who documents his process, and the video reflects a year of practical AI integration rather than a demonstration of new tools at launch.

The video is aimed at practising UX designers who want to see a real workflow in use, not a capabilities overview. Wooldridge is explicit early on that the goal is not to show every available tool but to show how specific tools solved specific problems in actual client work.

Key takeaways:

1. FigJam AI for workshop facilitation. Wooldridge describes using FigJam’s AI features to generate workshop templates, mind maps, and diagram structures from conversations. He notes it is not a breakthrough but removes the blank-canvas problem at the start of workshops, allowing more session time for actual facilitation. The ChatGPT plugin for turning discussions into flow charts also gets a mention as a tool that delivers consistent, if modest, results.

2. Prompt-driven wireframing. Wooldridge uses Figma Make to generate initial wireframe structures from written briefs. He shows examples of prompts that work well (specific constraints, named components, explicit reference to the design system) and prompts that produce unusable output (vague descriptions, no layout guidance). His position is that AI gets you past the blank screen and into the refinement stage faster, but the refinement stage is where the real design thinking happens.

3. AI for design QA. One of the most practically useful sections covers using a structured AI prompt to audit Figma frames against a design system. Wooldridge describes defining spacing scales and variable rules, creating a reusable prompt that checks for consistency, and running it across a file. The AI identifies violations that a manual review would catch, but faster and without the attention fatigue of checking hundreds of frames by hand.

4. Where AI still falls short. Wooldridge is direct that AI-generated layouts lack the subtle compositional judgment that comes from understanding a user’s context. He keeps a clear line between AI handling structural and repetitive tasks and the designer making decisions about hierarchy, emphasis, and emotional tone.

5. Tool selection principle. Rather than maintaining a large stack, Wooldridge advocates picking one AI tool per task category and learning it well. Spreading attention across many tools, in his experience, means mastering none of them.

Worth watching if you are a UX designer looking for a grounded account of what AI actually changes in day-to-day work in 2026, as distinct from what the tools claim to change.