Deep dive — chart 1

Chart 1 — rolling mean underwriting risk-score by borrower group

The chart your Chief Compliance Officer should look at first. Where the silent B3 divergence begins, in one visual.

7-day rolling mean underwriting risk-score per borrower group over 90 days

7-day rolling mean underwriting risk-score per borrower group over the 90-day audit period. B3 divergence begins Day 44 and never recovers.

What you are looking at

X-axis: Day 1 through Day 90 of the audit period
Y-axis: 7-day rolling mean of the AI's underwriting risk-score (0-100 scale, higher = riskier)
Each line: one borrower group (B1 = prime higher-income, B2 = near-prime mixed, B3 = age 25-45 near-prime mixed lower-income zip, B4 = sub-prime thin file, B5 = sub-prime derogatory)
Baseline window: the first 30 days establishes what "normal" looks like per borrower group

What the eye sees

Four borrower groups stay roughly parallel across the 90 days. One borrower group — B3 — drifts up. The rise begins around Day 44, and it does not recover. The line separates from the pack — upward, into higher-risk territory.

This is the pattern a Fair Lending Officer instinctively recognizes: something in the B3 flow changed, and the AI is now scoring that flow as much riskier than baseline. The 44 percentage-point lane-shift follow-on is the operational consequence.

What the sensor fires on

The sensor is not looking at the eye's pattern. It is running four quantitative thresholds against each 7-day window per borrower group:

Mean-shift: is the 7-day mean > 1.5 standard deviations from baseline mean?
Variance-ratio: is the 7-day variance > 1.5x baseline variance?
KL-divergence: is the shape divergence between baseline and current distribution > 0.2?
Group-differential: is the max cross-group shift > 12%?

Day 44 fires B3 KL-divergence 0.587 (nearly 3x the threshold). Day 86 hits 0.663. The pattern the eye sees is the pattern the sensor quantifies.

What fair-lending reviewers care about

Fair-lending reviewers do not read charts — they read attestations. But when an attestation is challenged, the chart is the exhibit. CFPB supervisory examiners, OCC / Fed / FDIC exam teams, DOJ pattern-or-practice investigators, state AG fair-lending desks, and plaintiff class counsel all understand the "one borrower group drifts silently" pattern instantly.

The chart also tells the counterfactual story: had this monitoring been in place operationally, the drift window would have been ~7-14 days rather than 45+.

The next chart

The rolling-mean chart shows when. The baseline-vs-drifted chart shows how the B3 risk-score distribution shape shifted.

Chart 2 — baseline vs recent distribution →

$499 Snapshot. 3 business days.

Same 3-chart set for your bank's actual AI credit-underwriting surface + full analysis report.

Buy $499

Back to banking sample hub.