Definition
The change in adjudication time and ruling accuracy on claims where no evidence package can be assembled, measured against the same adjudicators' performance on assisted claims and against their own earlier baseline.
The Measurement Gap
There is no metric that tells you how adjudication time and accuracy change on claims where the AI cannot assemble an evidence package, which means capability loss surfaces only in the hardest and most consequential cases.
Formula
UAC(t) = ( acc_unassisted(t) / acc_assisted(t) ) รท ( acc_unassisted(t0) / acc_assisted(t0) ) , reported with the time ratio T(t) = median_time_unassisted(t) / median_time_assisted(t)
Proposed formula, pending validation pilots.
At a Glance
Decision Supported
Whether the adjudication bench can still work the hardest class of claim, and whether unassisted practice needs to be scheduled.