# Select metrics
<a name="rairc03"></a>


| RAIRC03: How will you select metrics for the release criteria across each responsible AI dimension? | 
| --- | 
|   | 

 Consider specialized metrics for release criteria related to your identified potential harms.

**Topics**
+ [RAIRC03-BP01 Measure safety harms and harmful outputs](rairc03-bp01.md)
+ [RAIRC03-BP02 Measure fairness as unwanted bias across stakeholder groups](rairc03-bp02.md)
+ [RAIRC03-BP03 Measure veracity of outputs](rairc03-bp03.md)
+ [RAIRC03-BP04 Measure robustness of outputs to input variation](rairc03-bp04.md)
+ [RAIRC03-BP05 Measure privacy protection](rairc03-bp05.md)
+ [RAIRC03-BP07 Measure user controllability of system behavior](rairc03-bp07.md)
+ [RAIRC03-BP08 Measure explainability of system behavior](rairc03-bp08.md)
+ [RAIRC03-BP09 Measure security risks and threats](rairc03-bp09.md)
+ [RAIRC03-BP10 Measure transparency quality](rairc03-bp10.md)