PEP and sanctions screening with AI Analysis
Screen against 1,000+ sanctions, PEP and watchlist sources. Ask AI Analysis to explain a match; your team decides.
- 1,000+watchlists
- <1sresponse
- 24hrefresh
- 195countries
A sample screening console: a name check, scored matches and an AI Analysis answer.
Screening in production for
What every scan checks
Sanctions
OFAC, UN, EU and HMT.
PEPs
PEPs, relatives and close associates.
Watchlists
Law-enforcement, regulatory and interest lists.
Adverse media
Negative news, person or entity.
Data-breach exposure
Email breach history, supplementary.
Original-script names
Arabic, Cyrillic, Chinese, Korean, Japanese.
Fewer false matches, with matching you tune
Phonetic and fuzzy matching handle spelling and transliteration. Matching Scope sets how far DOB, gender, nationality and ID apply; blank-profile policies enforce or ignore them, and weak matches drop below threshold.
- Matching Scope per field
- Blank-profile scan policies
- Keyword search in results
A sample matching result: name variants scored against a threshold, with weak matches filtered out.
Ask the match. Get the reasoning
Ask AI Analysis about any matched profile, individual or corporate, in the app or via the API. It explains which identifiers align; your team decides.
- Enabled per organisation
- Single scans, batch and API
- Explains the match, never decides
A sample AI Analysis panel answering a question about a matched profile.
Daily re-screening, with alerts that arrive
Monitoring-list customers are re-screened daily, with a full rescan at renewal. New listings and profile changes raise alerts, and API callbacks push them into your systems.
- Re-screened daily while listed
- Full rescan at renewal
- Alerts and API callbacks
A sample monitoring timeline: daily re-screens with one alert on a new listing.
A case record a regulator can follow
Each match keeps its decision, risk level, comments, assignee and history.
- Decision, risk level and comments
- Assignee and full history
- Custom watchlist risk levels
A sample case record with a decision history and a recorded outcome.
From a thousand sources to one decision
The shares are illustrative, but the shape is real: each stage removes noise before the last one, where analyst time is spent.
- Your team
Configure
Thresholds, fuzzy logic and secondary identifiers, set to your risk appetite.
- Automated
Screen
Every customer against 1,000+ sanctions, PEP and watchlist sources.
- Automated
Candidate matches
Every name that touched a list, before any scoring is applied.
- Automated
Scored and filtered
Weak coincidental matches drop out below your threshold.
- Your team
Review and decide
Evidence, assignee and a decision history a regulator can inspect.
1,000+ sources behind every scan, one decision trail
- 1,000+
- Global watchlists screened
- <1s
- Average screening response
- 24h
- Watchlist refresh cycle
- 195
- Countries covered
Assurance
ISO/IEC 27001:2022
ISO/IEC 27701:2019
GDPR
DORA
Three ways to run a scan
Web app
Interface: English, Japanese, Spanish, French, Chinese, Arabic. Reports: French, Chinese, Arabic, Japanese.
Batch upload
CSV or XML; preview flags risky entries.
REST API
Screening inside your systems.
POST /api/v3/member-scans/single
{
"firstName": "Jonathan",
"lastName": "Sample",
"dob": "1971-04-12",
"nationality": "FR"
}
200 OK
{
"scanId": "scan_8f21c",
"matchedNumber": 1
}
Common questions
Can I tune matching to cut false positives?
Yes. You set the name match type, the close-match threshold and how far secondary identifiers such as date of birth, gender, nationality and ID number narrow the results, so weak coincidental matches are filtered while genuine hits still surface.
Can screening run in real time via API, or in batch?
Both. Screen one customer at onboarding through the API for a sub-second response, or upload the whole book as a CSV or XML batch file. Both paths use the same lists and the same matching logic.
Does MemberCheck re-screen customers after onboarding?
Yes. Entities you add to the monitoring list are re-screened daily, so a previously cleared customer is reviewed again when a new listing or a profile change affects them. Paused and removed entities stop being checked.
Can you exclude domestic PEP noise for a local customer base?
Yes. A PEP jurisdiction policy includes or excludes PEP and close-associate profiles by country, which is how firms filter domestic politically exposed people when their risk picture does not require them. The setting is configured by your Compliance Officer.
Does the check include data-breach exposure?
Yes. Enter an email address with the individual and the result adds a list of known data breaches for that address — supplementary intelligence, not a regulatory KYC requirement.
How many sources do you screen against, and how often are they updated?
MemberCheck screens against over 1,000 sanctions, PEP, and watchlist sources, refreshed on a 24-hour cycle, covering 195 countries. The full list of covered lists is visible to your team in the List Access screen.
What counts as a politically exposed person?
FATF Recommendation 12 splits PEPs three ways: foreign PEPs holding a prominent public function in another country, domestic PEPs holding one at home, and people entrusted with a prominent function by an international organisation. Jurisdictions apply the categories differently, so a customer in scope in one country may not be in another.
Do you screen relatives and close associates as well as the PEP?
Yes. A PEP's family members and close associates are screened as part of the same check, because they are the obvious route for holding assets at one remove. Former PEPs and Profiles of Interest carry the same related-risk logic after office.
Which sanctions regimes are covered?
Screening covers the major consolidated regimes — OFAC, the United Nations Security Council, the European Union and HM Treasury — alongside law-enforcement, regulatory-enforcement and national lists. Coverage by country is set out on the country coverage pages.
How do you handle non-Latin names and transliteration?
Close name matching uses phonetic algorithms and fuzzy search on Latin names, and an Original Script Name field takes Arabic, Cyrillic, Chinese, Korean or Japanese text directly, since transliterated names rarely match character for character. Secondary identifiers narrow what remains.
What should an audit trail contain for an inspection?
A regulator asks what was screened, what matched, what the analyst concluded and on what evidence. Each match keeps its source list, score, the identifiers used, its assignee and the decision that closed it, so the record answers the question without reconstruction.
Can different customer segments have different thresholds?
Yes. Settings can be preset per organisation or suborganisation and left user-defined for authorised staff, so the threshold follows the segment's risk instead of one average across the whole book. A single threshold either buries analysts in noise or sets the bar too high.
Further reading
Pairs with
See MemberCheck in action
A short walkthrough of screening, verification and ongoing monitoring in one platform, set to the thresholds and jurisdictions your programme actually uses.