COMPEL Glossary / quantitative-risk-assessment
Quantitative Risk Assessment
Quantitative risk assessment is an approach to evaluating AI risks that uses numerical data, statistical methods, and mathematical models to estimate the probability and potential financial or operational impact of identified risks.
What this means in practice
Unlike qualitative assessment (which uses categorical ratings like 'high/medium/low'), quantitative assessment produces specific numerical estimates: 'There is a 15% probability of this model producing biased outcomes, with an estimated financial impact of $2M in regulatory penalties.' Quantitative assessment requires more data and expertise than qualitative approaches but enables more precise risk-return tradeoff analysis, especially for high-impact AI systems. In the COMPEL framework, quantitative risk modeling supplements qualitative assessment for high-risk AI systems starting at maturity Level 3.5.
Why it matters
Quantitative risk assessment produces specific numerical estimates of probability and financial impact, enabling more precise risk-return tradeoff analysis than qualitative high/medium/low ratings. For high-impact AI systems where governance decisions involve significant financial stakes, quantitative assessment supports evidence-based decisions about risk acceptance, mitigation investment, and insurance procurement. It transforms risk management from subjective judgment to data-driven analysis.
How COMPEL uses it
In COMPEL, quantitative risk modeling supplements qualitative assessment for high-risk AI systems starting at maturity Level 3.5. During the Model stage, the Governance pillar designs risk assessment methodologies appropriate to system risk classification. The Evaluate stage uses quantitative risk data for portfolio-level risk aggregation, and the Learn stage refines probability estimates based on observed outcomes across COMPEL cycles.
Related Terms
Other glossary terms mentioned in this entry's definition and context.