AI Glossary
Key Risk Indicator (KRI)
A measurable metric used to monitor changes in risk over time.
Key Risk Indicator (KRI)
Overview
Organizations need a way to determine whether risk is increasing, decreasing, or remaining stable.
One way they do this is by tracking Key Risk Indicators, commonly known as KRIs.
A Key Risk Indicator is a measurable metric used to monitor changes in organizational risk and provide early warning signs of potential issues.
A helpful way to think about a KRI is the warning lights on a vehicle dashboard.
The lights do not fix the problem.
They alert the driver before a small issue becomes a larger one.
Organizations use KRIs in a similar way.
Examples of AI-related KRIs may include:
- Model accuracy trends
- Security incident frequency
- Number of compliance findings
- Data quality issues
- Failed model deployments
- Privacy incidents
- Bias monitoring results
Tracking KRIs helps organizations identify emerging risks, prioritize responses, and improve governance.
As AI programs mature, KRIs become an increasingly valuable part of continuous monitoring and risk management.
Why It Matters
KRIs provide early visibility into changing AI risks and support proactive decision-making.
Real-World Example
A company monitors the monthly number of AI security incidents and investigates any significant increase before it affects operations.