When the evidence is insufficient, the system should not invent a finding.
We used a public online-casino payments dataset as a discipline check. The source contains deposits and withdrawals, but not the full customer-interaction evidence required for the current safer-gambling assurance workflow.
Public Mendeley dataset
Dataset DOI: 10.17632/9j5gcygnwg.1. The publisher describes the casino file as 138,928 transaction observations, five columns and 2,286 customers.
Compatibility Report · 0 findings
The historical validation run recorded no risk findings from the payments-only file. That result is treated as an evidence-boundary check, not as proof of detection quality.
No evidence is not a PASS.
Missing required evidence must not be converted into a stronger conclusion. A compatibility result is different from a regulatory assurance result.
Not a certification. Not a passed audit. Not an operator benchmark.
This validation demonstrates evidence discipline only. Current GB customer-interaction assurance must be assessed on evidence relevant to SR Code 3.4.3.
View the public dataset →