From Law to Interface

A designer's guide to the EU AI Act

Guidelines alone are abstract. This is an attempt to make them buildable.

The EU AI Act sets out clear obligations for anyone building or deploying AI systems in Europe. But it was written by regulators, not designers. Translating legal requirements into components, patterns, and checklists — that translation work is largely missing.

This guide is my attempt to fill that gap.

I am a designer, not a legal professional. I read the Act, cross-referenced it with existing AI UX research including the Shape of AI, Microsoft HAX Guidelines, and my own experience designing AI guidelines and components for a large pharmaceutical company, and asked a simple question: what does each obligation actually look like in an interface?

Some of the patterns already exist. They just have not been connected to the legal obligations they fulfill. Others are genuinely missing — gaps in the design community's collective thinking that the Act is now forcing us to address. I will keep adding to this collection as the space evolves.

This is written from a practical, design-first point of view. It is not legal advice. If you are assessing compliance for a high-risk AI system, involve a legal professional.

But if you are a designer trying to understand what you need to build — this is for you.

If you work in AI product design and want to contribute, I would love to hear from you.