A common surname collides with the list

Normalization

• Beneficiary name (field 59): “OMAR K. HADDAD” → “OMAR K HADDAD” (Strip punctuation; the middle initial is kept as a separate token but carries little weight on its own.)
• List entry primary name: “HADDAD, Omar” → “OMAR HADDAD” (Resolve the surname-first inversion so list names and payment names are compared in the same token order.)

Candidates

• le-haddad-orion: matched on Full-name match after inversion and middle-initial handling — Both name tokens of the list entry appear in the beneficiary name in comparable order; the alias set (Hadad, Umar) confirms the entry is indexed under common spelling variants.

Scores

• le-haddad-orion: total 0.83 vs threshold 0.82 (Name similarity 55%×96%; Phonetic similarity 20%×90%; Secondary identifier corroboration 15%×50%; Country corroboration 10%×50%)

Disposition

Alert generated. The total of 0.83 meets the 0.82 threshold on the strength of the name alone. The message itself contains nothing that can safely discount the match, so it must go to a human with access to more data than the message carries.

Investigation

1. Is the beneficiary the same individual as list entry le-haddad-orion? Evidence: Beneficiary bank KYC response: account holder date of birth 1988-09-02; List entry date of birth: 1961-04-17. Finding: A 27-year gap between the two dates of birth makes identity with the listed person implausible.
2. Does geography corroborate or contradict the match? Evidence: KYC file shows Canadian residency since 2015 with a Toronto utility bill on record; List entry records a last known location in Cyprus with no Canadian connection. Finding: There is no geographic overlap between the beneficiary and the listed person.
3. Do any hard identifiers overlap? Evidence: The passport number in the beneficiary's KYC file differs from the fictional passport K0448127 recorded on the list entry; No alias, former name, or transliteration variant in the KYC file links to the entry. Finding: No identifier connects the beneficiary to the list entry; the collision is purely a common name.

Final: False positive — release. Release the payment and record which identifiers discounted the match. Register the beneficiary-to-entry pairing so an identical future hit can be adjudicated from the recorded rationale — with the pairing re-reviewed whenever the list entry changes. Institutions differ on whether such pairings suppress future alerts automatically or only pre-populate the analyst's view.