TL;DR: AI won't design your system's taste, but it removes a huge amount of the busywork around one: naming tokens, writing component docs, auditing for inconsistencies, and drafting usage guidelines. Use it for the scaffolding so you can spend your time on the decisions that matter.

Can AI build a design system?

Not the soul of one — the visual language and principles are yours. But a design system is also a mountain of documentation and consistency work, and that's exactly where AI earns its keep.

1. Name and organize design tokens

Ask AI to propose a consistent naming scheme for colors, spacing, and typography tokens, and to flag where your current names are ambiguous or overlapping.

2. Draft component documentation

Give it a component and its props, and have it draft usage guidelines, do's and don'ts, and accessibility notes. You edit for accuracy instead of writing from a blank page.

3. Audit for inconsistency

Feed it a list of your styles or components and ask where things have drifted — near-duplicate colors, one-off spacing values, components that should be merged.

4. Generate variant ideas

For a new component, ask for the states and variants you might be missing (loading, empty, error, disabled). It's a fast completeness check before you ship.

FAQ

Should AI decide my design tokens' values? No — the values express your brand. Let it help with structure and naming, not taste.

Can it keep the system consistent over time? It's a strong auditing assistant, but a human still owns the system's direction.

Carlos Lastres is an Apple Design Award–winning product designer and software engineer based in Tokyo who works hands-on with AI tools to design conversion-focused products.