Generative UI

Generative UI refers to interfaces that are created or modified dynamically by AI systems in response to user context, intent, and history, rather than following a fixed design defined in advance.

What is generative UI in UX design?

Generative UI refers to interfaces that are assembled or modified in real time by AI systems rather than being defined statically in advance by designers. Instead of every user seeing the same predefined layout, a generative UI system uses information about the user's context, intent, history, and current task to compose an interface suited to that specific moment. A banking app might generate a bespoke micro-interface for a user who needs to dispute a specific charge, showing only the relevant transaction and a dispute button, rather than routing the user through a generic support flow designed for all possible issues.

How does generative UI change the role of UX designers?

In generative UI systems, designers shift from defining specific layouts and components to defining the rules, constraints, and design tokens that AI systems use to assemble interfaces. Rather than designing a screen, a designer defines what kinds of components can appear in what contexts, what rules govern their combination, what visual constraints maintain consistency, and what boundaries prevent the AI from producing experiences that violate design principles or accessibility requirements. This is a significant shift from designing artifacts to designing systems that generate artifacts.

What are the UX challenges of generative UI?

Generative UI creates new challenges around predictability and user trust. When the interface changes based on AI inference rather than user action, users may feel disoriented if they cannot predict or understand why the interface looks different from one session to the next. The mental model users build of a product depends on consistency: if the interface is different every time, users cannot build reliable expectations. Generative UI systems must balance personalization with predictability, and must be transparent about why the interface has been adapted when that adaptation might be surprising.

Related terms

Related guides