Talk to usTry the demo

Sanctions Screening Explained: The Lists, the Matching and the Review

KYC, AML and digital identification

For: Compliance analyst who runs or reviews sanctions screening

How the main sanctions lists are published, how fuzzy name matching and thresholds shape the alerts you review, and what a review step should record.

You work in compliance and you review the alerts that a sanctions screening process produces. You need three things explained plainly: who publishes the main sanctions lists and how they change, how name matching and thresholds turn a list into alerts and false positives, and what a good review step records. This article uses the publishers' own pages and the EU AML Regulation, and ends with a checklist you can apply to any screening setup.

Who publishes the lists

Three public publishers are described here, and their pages show how different the formats are.

The United States. The Office of Foreign Assets Control publishes its lists through the Sanctions List Service. Users can download the Specially Designated Nationals list, which contains the names of individuals, entities and groups designated by the office together with maritime vessels and aircraft that it blocks, or the consolidated list of non-SDN sanctions lists, which sit in separate data files that comply with the office's data standards. The service also offers custom datasets built from selected lists or programs, an archive of published delta files by year, and a search application.

The European Union. The European Commission's overview explains that the Union has over 40 different sanctions regimes in place, some mandated by the United Nations Security Council and others adopted autonomously by the Union. Decisions on adoption, renewal or lifting are taken by the Council of the European Union, and a regime becomes applicable law when published in the Official Journal. The measures include arms embargoes, travel bans, asset freezes and other economic restrictions. The obligations are binding on EU nationals, on people located in the Union and on people doing business there, and investigation of possible breaches falls to Member States and their competent authorities.

The United Kingdom. Since 28 January 2026, the UK Sanctions List, published by the Foreign, Commonwealth and Development Office, is the only source for UK sanctions designations. The older consolidated list of asset freeze targets, which was published for the Treasury by its sanctions implementation office, is no longer updated. The government's transition guidance tells businesses that their systems need to use data from the UK Sanctions List. The UK Sanctions List covers designations made under the Sanctions and Anti-Money Laundering Act 2018, whose restrictions can include financial, immigration, trade or transport sanctions.

The United Nations Security Council also maintains a consolidated list. This article does not describe it in detail because its public page could not be retrieved for this review; the EU overview above notes that some Union regimes are mandated by the Security Council.

Lists are moving data

A list is not a document you read once. The UK page keeps an update log, and in October 2026 it records changes on nine separate days up to the 26th, including new designations, variations, corrections and delistings under several regimes. The OFAC service keeps an archive of published delta files, which lets consumers see what changed between versions. For a screening setup, this means freshness is a measurable property: how soon after a publisher updates does your process screen against the new data, and who sees the delta?

From list to alert: name matching

A sanctions list contains names, aliases, dates and other attributes. A customer's name will rarely match a list entry character by character, because of transliteration, abbreviation, word order and typing errors. Screening therefore uses approximate, or fuzzy, matching with a threshold. OFAC's own search tool shows the mechanics in public.

The tool provides exact matches when the match threshold slider is set to 100 percent, and can provide a broader set of results using fuzzy logic that combines character and string matching with phonetic matching; only the name field invokes fuzzy logic. The score indicates the similarity between the name entered and a result on one of the lists. It is calculated using two matching logic algorithms, one based on phonetics and one on the similarity of the characters in the two strings, and the slider defaults to 100, which indicates an exact match; lower scores indicate potential matches. The search first looks for potential matches based on the first letter of the input terms and on matches at least 50% or more similar by edit distance. It then uses the Jaro-Winkler string difference algorithm and the Soundex phonetic algorithm, applies them both to the whole name string and to the individual name parts, and returns the higher of the two scores. The minimum name score field limits the number of names returned. A value of 100 returns only exact matches on the characters entered, a value of 50 returns all names deemed 50% similar, and lowering the threshold gives a broader result set.

The consequences follow directly. A lower threshold returns more candidates, so more of them are false positives that an analyst has to clear. A higher threshold returns fewer candidates and raises the chance of missing a variant spelling. There is no neutral setting, and the setting is a risk decision that someone must own and document. A screening policy should state the threshold, the matching methods, the lists and the fields used (name, date of birth, nationality), and the reasons.

Ownership and control under the new EU Regulation

The sections above describe sanctions lists and tools as the publishers' own pages present them today. The duties described from here on come from Regulation (EU) 2024/1624, which was published in the Official Journal on 19 June 2024 and, according to its final provision, applies from 10 July 2027, with a later date for certain categories of obliged entity listed in the text. They are upcoming duties and should be kept apart from the sanctions obligations that apply under each jurisdiction's law now.

Screening a person's name is not enough when the customer is a company. Article 20(1)(d) of the EU AML Regulation, Regulation (EU) 2024/1624, lists among the customer due diligence measures "verifying whether the customer or the beneficial owners are subject to targeted financial sanctions". For a legal entity, the same point covers whether natural or legal persons subject to targeted financial sanctions control the entity or have more than 50 % of its proprietary rights or a majority interest, whether individually or collectively. Screening has to follow the ownership structure, not just the customer's name.

Keeping screening current

Article 26(4) of the same Regulation adds that obliged entities shall regularly verify whether those conditions are met. The frequency has to be commensurate with the exposure of the obliged entity and the business relationship to the risk of non-implementation and evasion of targeted financial sanctions, and for credit institutions and financial institutions the verification is also carried out upon any new designation. The provision does not replace the obligation to apply targeted financial sanctions or stricter requirements under other Union or national law.

What the review step should record

Article 21(3) requires obliged entities to keep a record of the actions taken to comply with customer due diligence, including records of the decisions taken and the relevant supporting documents and justifications. The Regulation applies this to refusals and terminations too. For a sanctions alert, a record that a later reviewer can follow typically contains these elements, drawn from that requirement and from the matching facts above:

  • the list and list version or date that produced the alert;
  • the matched name, the screened name and the score or method;
  • the data points used to compare (date of birth, country, identifiers) and the result of each comparison;
  • the decision (clear, escalate, block), who took it and when;
  • the justification and the supporting documents.

A checklist for any screening setup

  1. Which lists are screened, from which publisher feed, and how often are they refreshed?
  2. Which fields are compared, and how are aliases, transliteration and word order handled?
  3. What is the threshold, who set it, and when was it last reviewed?
  4. Is ownership screened for companies, including indirect control?
  5. Are existing customers re-screened when a list changes?
  6. Can you reproduce, months later, why an alert was cleared?

How NeuroVision fits

NeuroVision offers international AML screening through a licensed data provider. It covers sanctions lists of OFAC, the EU, the UK and the UN and 50+ national lists, PEP screening that includes relatives and associates, and adverse media. Matching uses machine learning with transliteration, a threshold from 0 to 100 with a recommended value of 70, and two modes, Search and Monitor.

To see the screening options in your own scenario, contact the team.

Sources

All sources were fetched on 27 October 2026.

Ready to talk?

Tell us what you are building and we will help you plan the integration.

Talk to us