Integrity & financial-crime diligence
A screen that never implies clean
Screening tools answer hit or no-hit. That answer hides the question a compliance team actually needs settled: checked against what, and how fresh was it? Enakt answers both, in one document — and says plainly when it could not.
Three answers, because two are not enough
The catastrophic error in screening is the opposite of the one in financial diligence. There, the danger is a confidently wrong number. Here, it is a confidently clean miss — telling a team that a designated party is safe to onboard. So the result grammar is built around one rule: never let “not checked” collapse into “clean.”
Match found
The subject matches a designation on a primary list, and the match resolves to that list’s own entry id, version and retrieval time — with a written justification, never a bare similarity score.
No adverse record found
In the sources that were actually checked, as of a stated time. That is a claim about scope, not a clearance — and it is only ever read alongside the coverage manifest.
Not reached
A source that was unavailable, or that we deliberately did not run. A statement about our coverage, never about the subject.
What comes back
A section of a screening report, exactly as the product writes it. Each line is one source, its disposition, and — when something matched — the primary record it resolves to.
1 designation match found in the primary list.
Exact name match: subject 'ACME Holdings LLC' matches the primary name of OFAC_SDN entry SDN-10001 ('ACME HOLDINGS LLC') — normalized token sets identical.
No adverse record found in the sources checked, as of 2026-07-24.
No sources could be checked — this is a coverage gap, not a clean result.
The list could not be retrieved on this run, so nothing about it is claimed.
No adverse record found in the sources checked, as of 2026-07-24.
Dockets retrieved for this query that did not name the subject as a party are not shown.
Subject and match are illustrative fixtures, so no real party is named. The wording, structure and provenance fields are the product’s own output.
The coverage manifest ships with the result, not under it
Every report carries a machine-readable list of what was queried, how fresh that data was, and what was not reached — as a first-class part of the document rather than a methodology footnote. A regulator reviewing your file, or a partner countersigning it, can see the boundary of the screen without asking you for it.
The whole report is then sealed with a content hash covering every field — the subject, the dispositions, the coverage, and the quoted media. A copy that has been altered will not reproduce its own hash, so a document you forward months later is verifiable by whoever receives it.
Adverse media, without the guesswork
Media screening is where a tool is most tempted to assert something. Enakt does not. An article attaches only when the quoted sentence itself names the subject and carries language from a published list of adverse terms, and the quote is lifted verbatim from the page — not from a search snippet you cannot check by following the link.
Every such item sits in a labelled sourced, not verified tier. It carries the publisher, the date and the link; it states that the characterisation is the publisher’s and not ours; and it is structurally unable to change the subject’s disposition. An article that names the subject in its headline and the allegation three paragraphs down is dropped, because attributing that allegation would be an inference rather than a citation.
What is checked
- Sanctions designations
- Official OFAC SDN and Consolidated, EU Consolidated, and UN Security Council lists — fetched from the issuing body, not resold by a vendor.
- Federal court records
- US federal dockets via CourtListener’s RECAP archive, matched on named parties.
- Adverse media
- Published reporting, quoted verbatim, and only where the quoted sentence names the subject.
What is not
- Corporate registries and beneficial ownership
- Not checked. There is no primary registry source licensed for it yet, and a screen states this in every report rather than letting silence imply coverage.
- State and non-US court systems
- Not checked. RECAP is a crowd-sourced mirror of the federal PACER system; its coverage is not complete, so the absence of a docket is not evidence that no case exists.
- Anything behind a subject’s consent
- The screen uses public sources only, and collects no personal data beyond the identifier you give it.
This half of the page exists because a screen is only as good as its stated boundary. Anything not listed on the left is reported as a gap in every screen that runs — the product will not let a limit go unmentioned, and neither will this page.
One more boundary worth naming: the measured reliability figure Enakt publishes elsewhere on this site is scoped to financial values, periods and computations checked against primary SEC filings. It is not a screening accuracy claim. No equivalent measurement exists for this domain yet, so none is offered here — and you should treat any screening vendor’s unqualified accuracy percentage the same way.
Screen something you already know the answer to
That is the honest way to evaluate a screening tool: run a subject whose record you can already check, and see whether the report tells you what it did and did not do. Sign-up is open and takes a minute.