What is a false positive in sanctions screening?
The moment a screening tool shows you two people with the same name and one of them is on the SDN List, you are looking at a false-positive problem. A false positive is a candidate the system returned that, on review, turns out not to be the listed party.
False positives are the ordinary output of any configuration tuned to avoid misses. A list entry for a common Arabic, Chinese or Slavic name will match thousands of unrelated people, and entries carrying no date of birth cannot be separated from them by data alone. The mistake teams make is treating a high rate as a tuning problem solved by raising thresholds. That trades noise for false negatives, which is the failure a regulator actually penalises.
What this workflow covers
SCOPE- A candidate becomes a false positive once an analyst has compared identifiers and written down why it was dismissed.
- Two men named Ali Hassan with dates of birth fourteen years apart and different nationalities: dismissed, with that reasoning on file.
- Dismissals are not permanent. If the authority later adds a date of birth or an alias to the entry, the same pair is reviewed again rather than auto-suppressed.
- In Full Search each candidate arrives with readable reasoning and a false-positive flag instead of a bare score.
- Every dismissal is a decision that has to be attributable and dated.
Compliance glossary
TERMS- False negative
- A genuine list match the screening system did not return, usually because of a high threshold, missing alias handling or stale list data.
- Threshold
- The similarity level at which a screening engine reports a candidate; lowering it raises recall and noise, raising it raises the risk of misses.
- Disposition
- The recorded outcome of reviewing a candidate: true match, false positive, or escalate.
Authoritative references
SOURCES- 01A Framework for OFAC Compliance Commitments
U.S. Department of the Treasury — OFAC
- 02Financial sanctions general guidance
Office of Financial Sanctions Implementation, HM Treasury
Frequently asked questions
Q&A- What is actually causing our false positives?
- Common names, transliteration and spelling variants, partial name matches, missing identifiers on the list entry, and thresholds deliberately set low to avoid misses. Poor input data multiplies all of them, which is why the cheapest fix is usually upstream of the screening tool.
- How do we cut the noise without missing a true match?
- Improve the input first: legal names, country, date of birth, registration number. Then dismiss on secondary identifiers rather than by raising the match threshold, and carry previous dispositions forward so the same candidate does not come back until the underlying record changes.
- What is a normal false-positive rate?
- There is no published benchmark that transfers between firms, because it depends on the customer base, the name distribution and the threshold policy. Supervisors look at whether the rate is measured, explained and reviewed, and whether dismissals are documented.
- What is a false negative?
- A true match the system failed to return, which is the opposite and more dangerous failure. It comes from thresholds set too high, missing alias or transliteration handling, stale list data or incomplete list scope, and it is what produces an actual breach.