FollowTheMoney JSON

Our JSON files follow the FollowTheMoney entity layout: each entity has an id, a schema and a set of properties whose values are always lists of strings. They are written as JSON lines (one complete JSON object per line), so they can be streamed without loading the whole file.

Two files

entities.ftm.json holds every record as an entity, followed on the next line by its Sanction entity: the listing on the source register. There are two lines per record. Use it to load a graph or an entity store, where the listing is a separate node.
targets.nested.json holds one line per record, with its listing nested under properties.sanctions. Use it when you want one self-contained object per listed party.

Both are UTF-8. See also the model reference, entity structure and flat and nested entities.

Entity fields

id: Record id, crown-<number>. One record is one row from one source register. The listing has the id crown-<number>-sanction. In the API, the same record is the number alone.
caption: Display name.
schema: Person, Company, Organization, Vessel, Airplane, LegalEntity (type not stated by the source) or Sanction.
properties: Property name to a list of string values. Absent properties are left out.
referents: Other ids merged into this one. Always an empty list today; merges are expressed with canonical_id.
datasets: Source register as a short identifier, e.g. ofac_sdn, uk_ofsi, paris_mou_banned.
first_seen: When we first collected the record (ISO 8601, UTC offset included).
last_seen: When the record was last confirmed at its source.
last_change: When the record last changed. Today this equals last_seen.
target: true on listed parties. Not present on Sanction entities.
canonical_id: Present when we have matched the record with records from other registers as the same party: crown-e-<merge id>. Group on it to see one party across registers.

Properties

Property	On	Contents
`name`	all records	Primary name as published
`alias`	all records	Other names and spellings
`topics`	all records	One of `sanction`, `role.pep`, `debarment`, `reg.action`
`birthDate`	Person	`YYYY`, `YYYY-MM` or `YYYY-MM-DD`, to the precision published
`nationality`	Person	ISO 3166-1 alpha-2, lower case
`country`	other records	ISO 3166-1 alpha-2, lower case
`imoNumber`	Vessel and others	`IMO` followed by seven digits
`sourceUrl`	records, Sanction	Link to the record or list at the publisher
`notes`	all records	Other published details as text (addresses, identifiers, flags, remarks)
`entity`	Sanction	id of the listed record
`authority`	Sanction	Source register name
`program`	Sanction	Programme or listing basis, in the publisher's words
`listingDate`	Sanction	Publication or listing date, where the register gives one
`sanctions`	nested file only	The record's Sanction entities

How complete each property is

Collection

Property	Records with it	Share
Loading…

Example: entities.ftm.json

Two lines from the maritime collection: a vessel and its listing (line breaks added for reading; each object is one line in the file).

{"id":"crown-380054154","caption":"PELIKAN","schema":"Vessel",
 "properties":{"name":["PELIKAN"],"topics":["debarment"],"alias":["BALTIC ATLANT SHPG-LITHUANIA"],
   "imoNumber":["IMO6922169"],"notes":["IMO 6922169 · Flag: Comoros · From: 09/12/2009 · ISM: BALTIC ATLANT SHPG-LITHUANIA"]},
 "referents":[],"datasets":["paris_mou_banned"],"first_seen":"2026-09-24T11:38:49.104591+00:00",
 "last_seen":"2026-09-24T11:38:49.104591+00:00","last_change":"2026-09-24T11:38:49.104591+00:00","target":true}
{"id":"crown-380054154-sanction","schema":"Sanction","caption":"Banned from Paris MoU ports — Multiple detentions",
 "properties":{"entity":["crown-380054154"],"authority":["Paris MoU Banned"],
   "program":["Banned from Paris MoU ports — Multiple detentions"]},
 "datasets":["paris_mou_banned"],"first_seen":"2026-09-24T11:38:49.104591+00:00",
 "last_seen":"2026-09-24T11:38:49.104591+00:00","last_change":"2026-09-24T11:38:49.104591+00:00"}

Example: targets.nested.json

The same record in the nested file: one line, with the listing inside properties.sanctions.

{"id":"crown-380054154","caption":"PELIKAN","schema":"Vessel",
 "properties":{"name":["PELIKAN"],"topics":["debarment"],"alias":["BALTIC ATLANT SHPG-LITHUANIA"],
   "imoNumber":["IMO6922169"],"notes":["…"],
   "sanctions":[{"id":"crown-380054154-sanction","schema":"Sanction",
     "caption":"Banned from Paris MoU ports — Multiple detentions",
     "properties":{"entity":["crown-380054154"],"authority":["Paris MoU Banned"],
       "program":["Banned from Paris MoU ports — Multiple detentions"]},
     "datasets":["paris_mou_banned"], "first_seen":"…","last_seen":"…","last_change":"…"}]},
 "referents":[],"datasets":["paris_mou_banned"],"first_seen":"…","last_seen":"…","last_change":"…","target":true}

Reading the files

Read one line at a time; the larger collections run to gigabytes (default: —).

import json

with open("targets.nested.json", encoding="utf-8") as f:
    for line in f:
        e = json.loads(line)
        names = e["properties"].get("name", []) + e["properties"].get("alias", [])
        programmes = [p for s in e["properties"].get("sanctions", [])
                      for p in s["properties"].get("program", [])]
        print(e["id"], e["schema"], names[0], programmes)

# group records that we have matched as the same party
jq -r 'select(.canonical_id) | [.canonical_id, .id, .caption] | @tsv' targets.nested.json | sort

What the files do not hold yet

Addresses, identity documents, ownership, family and associate links, positions held and crypto-wallet addresses are not separate entities or properties. Where a publisher gives them, they survive only as text in notes.
referents is always empty; merges appear only through canonical_id.
last_change is the last confirmation date, not a true change date.
Entity types are inferred where the source does not give one, and some are wrong; see methodology.

Last updated 2026-10-03