Sanctions Screening

Screen names against the OFAC SDN list, the OFAC consolidated list, the EU list and the UN list in one pass. Every row says which name matched, what the publisher thinks of that name, and which lists the answer is actually based on.

A list that did not download is not a clean answer

This is the rule the Actor is built around. The EU file is 26 MB and took 46 seconds on a good connection, so a run can lose it. When that happens the run does not quietly report no match. Every row carries listsSearched, listsUnavailable and screeningComplete, and a clean row with screeningComplete false means only that nothing was found on the lists that answered.

Weak aliases

OFAC marks each alias strong or weak, and the list published on 28 August 2026 carries 4,395 weak aliases out of 24,542. Weak is the publisher saying it expects false positives on that name. The UN does the same thing with Good and Low.

Those matches are returned, because a missed designation is worse than one you have to read. They are never presented as equal to a hit on the designated name: matchedOn says primary or alias, aliasQuality says what the publisher called it, and weakAliasMatch is a column you can filter on.

Not every publisher uses the field. The EU sets strong on all 31,053 of its aliases, so it draws no distinction at all, and EU rows report no quality rather than a strong that would look like OFAC's. The UN field also holds a.k.a. and f.k.a., which say nothing about quality, so those are left unknown with the original word kept.

The four lists designate different people

Screening only OFAC is not a screen for an EU obligation and the reverse is also true. That is why all four are on by default. Each row names the list it came from and the publication date printed on that file, so an answer can be audited months later.

How a name is matched

Word order is ignored, because the lists write AL-ZOMOR, Abboud Abdul Latif Hassan and people type Abboud Al-Zomor. Accents and punctuation are stripped on both sides.

100   the same name after normalising
90    every word you typed appears in the name
85    a one word query found as a whole word in the name

Below 85 you are looking at people who share a surname, which is why 85 is the default. whyItMatched gives the reason in words on every row.

What this is not

It is a search of published lists, not a compliance decision. It returns candidates with the evidence attached so a person can decide. It does not tell you whether a company is owned by somebody on a list, which is a separate question that no name search answers.

Access

Free, no key, no login. All four lists are whole-file downloads rather than search APIs, so the Actor reads 20,000 records to answer a name and parses only the records that share a word with it. Screening two names against OFAC and the UN read 20,332 records in 225 milliseconds.