Skip to main content

AI UI Text

AI UI text is the primary means of communication between users and AI systems. It conveys information, instructions, errors, and feedback — ensuring users can understand and effectively interact with AI functionality.

These guidelines build on the existing UI Text Guidelines for AI-specific contexts, rather than replacing them. They are based on US English, the primary source language at SAP.

Understanding Users and Context​

Creating effective UI text for AI requires understanding who your users are and how they interact with the system:

  • Recognize preferences — tailor UI text to user habits.
  • Understand user goals — clarify what users aim to achieve.
  • Map user tasks — identify the specific tasks users will perform.
  • Determine the interaction context — consider devices and environments (mobile, web, on-the-go).
  • Adapt to the context — use short prompts for mobile, detailed instructions for web.

Types of UI Text for AI​

Explanations​

Make AI actions and decisions clear to users, enhancing transparency and trust. Keep explanations as concise as possible.

AI explanation text example

Do: "Created with AI. Consider verifying results."

Don't: "Generated utilizing advanced artificial intelligence algorithms. Validate output integrity prior to deployment."

Instructions​

Guide users through tasks within AI systems with step-by-step directions to minimize errors and confusion.

AI instruction text example 1

AI instruction text example 2

Do:

  • Offer step-by-step, easy-to-follow instructions. E.g. "Fix spelling and grammar", "Tell AI what to do".
  • Ensure users can efficiently use AI features.
  • Minimize user errors by providing clear and concise instructions.

Don't:

  • Don't give ambiguous instructions. E.g. "Make corrections".
  • Don't provide overly complex instructions. E.g. "Ask Artificial Intelligence to generate content".

Description​

Introduce users to AI capabilities and explain what the AI system does and how it operates.

AI description text example

Do:

  • ✅ Output may contain errors.
  • ✅ Analyzing request…

Don't:

  • ❌ Consider verifying AI output errors.
  • ❌ Artificial Intelligence is analyzing the current request…

Do:

  • Clearly describe AI functions and capabilities.
  • Let users know what to expect: "Output may contain errors."
  • Explain how users can engage with AI.

Don't:

  • Don't use generic or unclear descriptions.
  • Don't expect the user to trust AI output without verification.

Error Messages​

Provide clear, specific guidance on resolving issues. Help users understand what went wrong and offer actionable steps.

AI error message example

Do:

  • ✅ Something went wrong during the generation process. Please try again.
  • ✅ It looks like an error occurred during the generation process. Try generating again.
  • ✅ Your request couldn't be generated. Try refining your request and generating again.

Don't:

  • ❌ System error. Contact administrator.
  • ❌ Artificial intelligence is having trouble understanding you.

Do:

  • Help users resolve issues that arise during AI interactions.
  • Give clear and actionable directions on what went wrong and how to fix it.

Don't:

  • Don't provide vague or generic messages.
  • Don't suggest actions that are incorrect in the user's context.

Summary​

Designing UI text for AI interfaces involves clear, concise, user-friendly communication. Key principles:

  • Understand the target audience and their goals.
  • Use simple and direct language.
  • Maintain consistency in tone and terminology.
  • Provide immediate and clear feedback.
  • Consider accessibility — ensure text is readable for all users.