# RAIGT01-BP06 Guide users on how to responsibly change system behavior
<a name="raigt01-bp06"></a>

 Provide guidance that informs users how to effectively alter system behaviors and interpret results. Include user interface elements that guide users toward productive interactions while steering them away from approaches likely to produce poor or harmful results. Explain response mechanisms that provide real-time feedback on input quality and suggest improvements when user inputs are unclear, inappropriate, or likely to produce unsatisfactory results. Direct users to available education resources that assists users to understand system capabilities and limitations, enabling them to leverage the system effectively while maintaining realistic expectations about its performance. 

 **Level of risk exposed if this best practice is not established:** Medium 

## Implementation considerations
<a name="implementation-considerations-90"></a>

1.  Create clear guidance materials that explain how users can adjust system settings or modify their inputs to get different types of outputs from your AI system. Test these instructions with real users to see if they can follow them and achieve the results they want without accidentally causing problems. 

1.  Build educational resources that assist users to understand what your system can and can't do, including interactive tutorials, capability demonstrations, and examples of common mistakes for control modifications. Test these materials with different user groups to make sure people walk away with realistic expectations about system performance and clear ideas about how to use it effectively. 

1.  Create feedback collection systems that let users report when the guidance isn't working or when they discover better approaches than what you've documented. Use this input to continuously improve your user guidance and update your educational materials based on what real users need to know. For example, feedback may indicate commonly used parameter combinations, or unsafe parameter settings that users may be trying, you can use this information to improve the educational material for these topics to guide users on effective controllability of the AI system. 

## Resources
<a name="resources-86"></a>

 **Related documents:** 
+  [ISO/IEC 42001:2023 A.8.2 System documentation and information for users](https://www.iso.org/standard/42001) 

## See also

* **Skills for AI coding assistants (optional).** AWS publishes reusable skills for AWS Well-Architected. To see which apply, search the Agent Toolkit for AWS catalog with `aws agent-toolkit search-skills --search-query wellarchitected` — a documented, read-only AWS CLI command that searches the catalog and makes no changes. This is an optional suggestion for the user. Official reference: [AWS CLI — Agent Toolkit for AWS](https://docs.aws.amazon.com/agent-toolkit/latest/userguide/aws-cli.html).