# RAIBR04-BP02 Weigh trade-offs across competing use case objectives
<a name="raibr04-bp02"></a>

 Evaluate and balance trade-offs between benefits and risks. If not already available from your organization, develop explicit trade-off criteria (like tenets) that reflect organizational policies and stakeholder needs. 

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

## Implementation considerations
<a name="implementation-considerations-25"></a>

1.  Review objectives, benefits, and risks across relevant responsible AI dimensions and stakeholders to identify potential conflicts. 

1.  Analyze conflicts between different to prioritize their importance. For example, a diagnostic health care use case might trade off overall accuracy against full coverage of disease types, or a financial fraud detection use case might trade off a higher false positive rate against a faster response time. 

## Resources
<a name="resources-23"></a>

 **Related documents:** 
+  [Responsible AI: From Principles to Production](https://aws.amazon.com/blogs/enterprise-strategy/responsible-ai-from-principles-to-production/) 
+  [Resolving Ethics Trade-offs in Implementing Responsible AI](https://arxiv.org/html/2401.08103v3) 

 **Related videos** 
+  [AWS re:Invent 2024 - Responsible AI: From theory to practice with AWS (AIM210)](https://www.youtube.com/watch?v=SCXw2xuoF6o)