How the data is collected, matched and refreshed

1. Collection

Crown ingests each register directly from the source publisher — the government department, regulator, court, self-regulatory body or international authority that issues it — rather than from a secondary aggregator. Each feed is registered individually, with its own route, format and licence position tracked against that specific publisher. This is what lets every record on this site carry a named source rather than a generic "government list" label.

Where a source publishes structured data (a CSV, XML or API feed), Crown ingests it directly. Where a source publishes only a web page or PDF, Crown extracts the same fields through a parsing pipeline built for that specific publisher's format, so that a change in one source's layout does not silently corrupt another's.

2. Matching and entity resolution

Once collected, records are normalized into a common schema — name, alternate names, entity type, program, nationality, date of birth where published, and the originating list — so that a name can be checked against every list at once instead of one register at a time.

Entity type (individual, company, vessel, aircraft or organization) is taken from the source where it states one. Where it does not, it is inferred: a date of birth points to an individual; a corporate suffix or registered-business language points to a company; vessel and aircraft markers such as an IMO or tail number point to those categories; and on PEP registers of office-holders, names without such markers are treated as individuals. Inference is sometimes wrong, and some sources record the wrong type themselves (a few companies on sanctions lists appear as persons); these are being corrected. Where nothing points either way, the record is left as unknown, shown as LegalEntity in the bulk files.

Records that plausibly refer to the same underlying person or entity across different lists are linked to a shared identifier, so that a search surfaces every list a name appears on, rather than requiring a separate search per source.

3. Refresh cadence

Sources are re-run on their own published cadence — from multiple times a day for fast-moving sanctions feeds to weekly or as-published for slower registries — and the compilation is reprocessed after each run. Every figure shown on this site and every record carries an "as of" date reflecting the last confirmed run against that source, so the currency of a number is never left implicit.

Because refresh cadence varies by publisher, the same figure can move at different speeds for different categories: sanctions listings typically move faster than corporate registry filings. Where a source publisher updates its own record between Crown's refresh runs, the source publisher's record is authoritative in the interim — see the verification notice on every record and on Attribution.

4. Honest limitations

Crown does not warrant that its compilation captures every record, every field or every update issued by every source publisher, or that matching between sources is free of error. Absence of a name from Crown's compilation is not evidence that no listing exists. A small number of fields are supplemented from internally researched data or from feeds obtained under arrangements that restrict further redistribution; any such field is labelled as such on the record rather than presented as if it came directly from the named publisher.

Crown does not create, originate, adjudicate, certify or verify any designation, listing, debarment, PEP status, corporate filing or beneficial ownership disclosure. Every such determination is made solely by the relevant source publisher, under its own authority, criteria and process. Before relying on a record for a compliance, hiring, lending, contracting, tenancy or other decision about a named person or entity, verify the current status directly against the source publisher's own published record.

5. Attribution and provenance

Every record shows the source publisher, the specific list it was drawn from, and the date it was last confirmed against that source. This is a deliberate design choice: a claim like "sanctions data" is not verifiable on its own, while "SEMA consolidated list, confirmed today" is. See the full Attribution and disclaimer notice and the Data licence for the attribution wording required when reusing a record.

6. What we publish

The compiled records are published through search, the screening API and the bulk files. The bulk files hold only the registers whose terms allow redistribution (see licence positions), and are rebuilt from the register as a batch. For politically exposed persons, see the PEP methodology.

7. Security

Crown maintains administrative, technical and physical safeguards appropriate to the sensitivity of the data it holds, including encryption of data in transit and at rest and access limited to personnel on a need-to-know basis. See the Privacy notice for how personal information appearing within compiled records is handled.