Calibration

The degree to which a model's predicted probabilities match the actual likelihood of outcomes, ensuring that a model's confidence reflects its true accuracy.

When an AI says it's 90% confident, it should actually be right about 90% of the time. A well-calibrated AI's confidence levels match its actual accuracy.

Well-calibrated AI systems support better governance because decision makers can appropriately weight AI recommendations. Your governance framework should require calibration assessment as part of model validation.