AI UX PlaygroundNewsletterJoin 2K+ AI designers and PMs on Substack. New teardowns, patterns, and prompts as they drop.

UX design

AI Error Handling UX

Design error handling patterns for AI features including error messages, recovery paths, and user guidance.

Prompt

Design error handling UX for [AI feature].

AI Feature Context:
- Feature: [name]
- AI capability: [LLM / image generation / code / etc.]
- Error types: [describe known errors]

Provide:

1. Error Taxonomy
   - Error categories (network, model, input, rate limit, etc.)
   - Error severity levels
   - User impact of each error
   - Frequency expectations

2. Error Message Design
   For each error type:
   - User-friendly message
   - Technical details (if needed)
   - What went wrong (in plain language)
   - Why it happened (if helpful)
   - What user can do

3. Visual Error States
   - Error indicators
   - Visual hierarchy
   - Color and iconography
   - Animation/motion
   - Placement and timing

4. Recovery Mechanisms
   - Retry options
   - Alternative approaches
   - Fallback behaviors
   - User actions available
   - Automatic recovery (if applicable)

5. Contextual Help
   - Helpful guidance
   - Examples of correct input
   - Tips to avoid errors
   - Links to documentation
   - Support contact (if needed)

6. Error Prevention
   - Input validation
   - Proactive warnings
   - Confirmation dialogs
   - Rate limit indicators
   - Usage guidance

7. Progressive Error Handling
   - First error: Simple message
   - Repeated errors: More guidance
   - Persistent errors: Escalation path
   - Error history: Learn from patterns

8. Trust & Transparency
   - How to maintain trust during errors
   - Transparency about AI limitations
   - Confidence indicators
   - What's happening behind scenes

9. Accessibility
   - Screen reader support
   - Keyboard navigation
   - Error announcement
   - Focus management

10. Testing Scenarios
    - Error scenarios to test
    - Edge cases
    - Recovery flows
    - User testing approach

Format as a comprehensive error handling UX specification.

How to use

  1. 01

    Replace [AI feature], [name], [LLM / image generation / code / etc.], and [describe known errors] with your specific details

  2. 02

    Add context before the prompt: Describe your AI feature and error types. Example: "Feature: Text generation. AI capability: LLM (GPT-4). Error types: Rate limits, timeouts, low confidence outputs, network errors."

  3. 03

    If you have existing error messages: Paste current error messages. Say "Current error messages: [paste messages]"

  4. 04

    If you have error scenarios: List known error scenarios. Say "Error scenarios: [list scenarios]"

  5. 05

    Paste the modified prompt into your preferred AI tool, like ChatGPT or Claude

  6. 06

    Review the error handling spec: Check error taxonomy, error message design, recovery mechanisms, and trust patterns

  7. 07

    Verify error messages: Ensure error messages are user-friendly and actionable

  8. 08

    Ask for specifics: Request "Focus on rate limit errors" or "Add more recovery mechanisms" or "Detail trust patterns"

  9. 09

    Export to your tool: Copy the error handling spec to Figma, Notion, or your design documentation

  10. 10

    Use for implementation: Apply the error handling patterns to implement user-friendly error handling

Tips

  • Include AI model details: Mention the AI model you're using (e.g., "GPT-4" or "Claude 3.5") so AI can provide model-specific error guidance
  • Specify error types: List known error types (e.g., "Rate limits, timeouts, low confidence") so AI can design appropriate error handling
  • Request user-friendly examples: Ask "Show example error messages for each error type" to see actual copy
  • For trust patterns: Ask "Suggest trust patterns for AI errors" to maintain user trust during errors
  • Save as template: Reuse the error handling spec structure for future AI features