Assess whether notices match the actual decisioning reasons (6cc3c2)
August 31, 2026
SITUATION Adverse-action notice principal-reason sample arrived with an underwriter chat that used coded language for HMDA data-quality manager. That is a Fair Lending Redlining and HMDA Data decision on notices match the actual in a credit-card issuer changing line-assignment logic.
DECISION HMDA data-quality manager in a credit-card issuer changing line-assignment logic must choose Remove access or reverse the item / Temporary compensating control / Approve a documented exception / Hold using adverse-action notice principal-reason sample after an underwriter chat that used coded language.
HYPOTHESES TO TEST 1. HMDA data-quality manager can defend Remove access or reverse the item from adverse-action notice principal-reason sample after an underwriter chat that used coded language in a Fair Lending challenge. 2. HMDA data-quality manager cannot defend Remove access or reverse the item from adverse-action notice principal-reason sample; Temporary compensating control is what the extract actually supports after an underwriter chat that used coded language. 3. An underwriter chat that used coded language never reached the population in adverse-action notice principal-reason sample — reopen intake, do not close notices match the actual. 4. Two facts in adverse-action notice principal-reason sample after an underwriter chat that used coded language conflict for HMDA data-quality manager; hold this Redlining and HMDA Data file.
ANALYSIS REQUIRED 1. Match the adverse-action language to the facts in adverse-action notice principal-reason sample. 2. Check HMDA coding and underwriting policy against notices match the actual. 3. Compare adverse-action notice principal-reason sample to similarly situated files, second-review notes, and reason codes after an underwriter chat that used coded language. 4. For this Fair Lending Redlining and HMDA Data file, read adverse-action notice principal-reason sample against an underwriter chat that used coded language and write the one fact that would move notices match the actual for HMDA data-quality manager.
RECOMMENDATION Choose Remove access or reverse the item / Temporary compensating control / Approve a documented exception / Hold on this Fair Lending / Redlining and HMDA Data packet (adverse-action notice principal-reason sample after an underwriter chat that used coded language). The follow-on Redlining and HMDA Data action is what HMDA data-quality manager does next: implement the option, assign an owner, and log the missing fact.
COMMAND RETURNS - Bottom-line Fair Lending option on notices match the actual, then the evidence in adverse-action notice principal-reason sample, then the action for HMDA data-quality manager - Hypothesis scorecard against adverse-action notice principal-reason sample: supported / rejected / untestable - Redlining and HMDA Data finding in adverse-action notice principal-reason sample that a second reviewer can re-perform - Missing page in adverse-action notice principal-reason sample after an underwriter chat that used coded language, if any
Explore more
More Fair Lending prompts
- Assess whether a special-purpose program is well designed or a pretext
- Assess whether pricing disparities are justified by legitimate factors
- Assess whether line assignments have a disparate impact the bank will defend
- Assess whether to pause a product pending a lookback (1e5bf7)
- Assess whether dealer overlays introduce prohibited steering (c7a01b)
Explore related decision areas
- Assess whether cyber controls claimed are actually in force (2b7bec)Insurance Underwriting
- Assess whether a generative-AI incident is a policy breach or a model defectAI Governance
- Assess whether prior-acts and notice issues make D&O unbindable as submittedInsurance Underwriting
See governed multi-model AI on your own prompt
Compare GPT-5, Claude, and Gemini side by side, with human review and a decision record built in.

