Home / Wiki / T / Transaction Screening

Transaction Screening

Compliance Risk and ControlsUpdated 16 Nov 2026

Transaction screening is the check a bank or payment provider runs on each individual payment, before it is executed or released, to find any link to a sanctioned person, entity, vessel or country. The names of the parties, the banks in the chain, addresses, country codes and the free-text fields of the payment message are compared with official sanctions lists. A payment that produces a possible match is held until a compliance analyst decides whether it can go ahead.

It is one half of sanctions screening; the other half is customer or name screening, which checks the client base at onboarding and whenever lists change. Transaction screening differs from transaction monitoring, which reviews patterns of activity after the fact to detect money laundering. Screening is a real-time gate on a single payment; monitoring is a retrospective analysis of behaviour. The term payment screening is often used for the same control. Software that performs payment screening is often called a sanctions filter, since every payment instruction must pass through it before release.

Screening obligations come from sanctions law rather than from anti-money laundering rules, although both are usually run by the same compliance function. The lists most commonly loaded into screening systems are:

  • The United States. The SDN list and other lists published by the Office of Foreign Assets Control (OFAC), which apply to US persons and to many payments cleared in US dollars.
  • The United Nations. The Security Council consolidated list, which member states implement through their own law.
  • The European Union. The consolidated list of persons, groups and entities subject to EU financial sanctions.
  • The United Kingdom. Designations under UK sanctions regimes, with financial sanctions implemented by HM Treasury's Office of Financial Sanctions Implementation.

Firms add local lists where they operate and often their own internal lists. Under the US 50 percent rule, an entity owned 50 percent or more by blocked persons is itself treated as blocked even if it is not named, so good screening relies on ownership data as well as list entries. The international standards behind these regimes are set out in the FATF Recommendations on targeted financial sanctions related to terrorism and proliferation.

02

What is screened

In a cross-border credit transfer, such as a SWIFT MT103 or its ISO 20022 equivalent, the screening engine reads the ordering customer, the beneficiary, their addresses, the ordering, intermediary and beneficiary banks and their identifier codes, and the remittance information. Country names and codes, city names linked to sanctioned regions, and port or vessel names in trade-related payments are also checked. The richer, structured data in ISO 20022 messages makes it easier to screen each element separately and reduces some false hits caused by unstructured text.

Rules on wire transfer transparency require originator and beneficiary information to travel with the payment, so that every bank in a correspondent banking chain can screen it. Removing or altering that information to avoid screening, known as wire stripping, has led to some of the largest sanctions penalties imposed on banks.

03

Matching and false positives

Exact matching would miss spelling variants, transliterations, reordered names and deliberate misspellings, so screening engines use fuzzy matching. Common techniques include phonetic algorithms, edit-distance scores, token reordering and lists of known aliases. Each potential hit receives a score, and anything above a threshold set by the firm produces an alert.

The cost of fuzzy matching is a high false positive rate: the large majority of alerts are common names or partial word matches with no real connection to a listed party. Firms manage this by tuning thresholds, maintaining approved-match lists for customers already cleared, and suppressing known non-risky terms. Every tuning change should be documented and tested, because a threshold set too high to save analyst time can let a true match through. Regulators expect the settings to follow a documented risk-based approach.

04

Alert handling and outcomes

An alerted payment stops straight-through processing and enters a review queue, usually inside a case management system. An analyst compares the payment data with the list entry, using dates of birth, nationality, addresses and other identifiers, and records an alert disposition. Because screening is a pre-release control, alerts must be cleared quickly, often within the payment's cut-off time.

A confirmed match leads to one of two outcomes under US rules. If a blocked person has an interest in the funds, the payment is blocked and the money is placed in a segregated blocked account until OFAC licenses its release. If the payment is prohibited but no blocked interest exists, for example a transfer linked to an embargoed country, it is rejected and returned to the sender. Both blocked and rejected transactions must be reported to OFAC, generally within ten business days. Other jurisdictions use comparable asset-freeze and reporting rules.

05

Weaknesses and evasion

Screening only catches what the data reveals. Front companies, nominee owners, vague payment references and transfers split across several intermediaries are standard sanctions evasion methods designed to pass list checks. For this reason screening is combined with customer due diligence, ownership research and behavioural monitoring rather than relied on alone.

Last updated 16 Nov 2026