Overview
How might we design multi-step forms with ai so people can trust and act on AI output?
When this pattern fits
- Ideal for lead generation tools, onboarding flows, and data collection applications where adaptive, AI-powered forms improve completion rates and user experience.
When to skip or lighten it
- Legal forms that must show every field for compliance.
- Two-field forms where steps add friction.
- Users who need to see the full schema up front for planning.
States
State model coming soon
Key UX elements
Key UX elements coming soon
Anti-patterns to avoid
Skipping steps without a way to revisit them.
Prefill that overwrites carefully typed user data.
No progress indicator across adaptive paths.
Asking the same question again after AI already inferred it.
How products use it
| Product | Implementation |
|---|---|
| Typeform | Logic jumps and AI-assisted branching in surveys. |
| Jotform | Conditional multi-page forms with smart defaults. |
| Google Forms | Branching sections; AI assists drafting questions. |
| SurveyMonkey | Skip logic and adaptive questionnaires. |
Implementation
Copy this prompt to generate a production-ready implementation in Cursor, Claude Code, Lovable, or any AI coding agent.
Generate a production-ready implementation of the "Multi-step Forms with AI" AI interface design pattern.
Pattern Definition:Frequently asked questions
What makes a multi-step form “AI”?
The path, defaults, or clarifying questions change based on model inference from earlier answers—not only static if/else rules.
Should users see skipped steps?
Offer a review or “show all fields” mode. Hidden forever steps frustrate people who need an edge-case field.
How does this differ from use-case wizards?
Wizards onboard to a product goal. Multi-step AI forms collect structured data for a submission, ticket, or application.
What about accessibility?
Announce step changes, keep focus management sane, and do not rely on animation alone to signal progress.