Definition
Measures the rate at which risk segments in the AI-underwritten book accumulate loss ratios deviating from model prediction, detecting systematic underpricing due to applicant or broker exploitation.
Formula
ASAR = SUM (NB_hot_segments / NB_total) x (Actual_LR / Predicted_LR)
Proposed formula, pending validation pilots.
At a Glance
Category
AI Underwriting & Claims Judgment
Decision Supported
Is the AI underwriting model being gamed by applicants or brokers who learned which risk profiles it underprices?