WritfindFindings, not flags.

How a report is built

  1. 1

    Screen public sources

    We search for adverse information within the investigation scope, screen the results and retrieve relevant pages. Any use of a search snippet is marked.

  2. 2

    Investigate and write by matter

    Related events are written together in one chapter. Where further research could change the report, we look for evidence about unresolved outcomes or conflicting accounts.

  3. 3

    Link findings to source passages

    Before a report is assembled, quoted passages are checked against the saved source text. A text match does not by itself prove that a claim is supported.

  4. 4

    Make the chronology visible

    Key dates from the report form a timeline. Each selected date leads back to its sentence and the matter it belongs to.

Reports distinguish events from the dates they were reported or recorded. Information is limited to the source-reading date shown in the report.

Reading the report’s status labels

Category results use Flagged, Clear and Undecided. Unavailable content and unread sources are marked separately.

  • FlaggedAdverse matters were found.
  • ClearThe category search was completed and no adverse matters were found.
  • UndecidedThe category has not been fully assessed.
  • Not surveyedThe relevant report content or source text is unavailable.

The nine categories

Reports show the search result or assessment status for each category.

  1. Fraud
  2. Securities violations
  3. Money laundering
  4. Sanctions
  5. Bribery and corruption
  6. Organised crime
  7. Tax evasion
  8. Terrorism financing
  9. Other adverse matters

How the benchmark scores a report

Writfind conducted this comparison in September 2026 on three development cases using the same investigation scope. Coverage and source support used model-assisted judgments with tool identities withheld. Human decisions were recorded for reference-set exceptions; this was not an independent human audit. The published comparisons include three graded Writfind report versions. Each Deep Research report was generated once; revisions and rescoring are not independent repeated trials.

Recall
Matters found divided by the admitted reference set. The set pools candidates from the compared reports and retains five matters from earlier research. It cannot count unknown matters missed by every investigation.
Precision
Matters found in the admitted reference set divided by that count plus recorded candidates rejected as not adverse, wrong-subject or out-of-scope. Writfind's two rejections were carried over from an earlier report version; this is a calculation from recorded counts, not a fresh precision measurement of its current report.
F1
The harmonic mean of recall and precision. It inherits the version limitation of the precision calculation.
Evidence support
Passages whose central claim is supported by a cited source, divided by the eligible cited passages. This does not mean every detail is supported or measure whole-report accuracy. Uncited passages, passages with no readable cited page and investigation-only passages are excluded. Unresolved judgments remain in the denominator.
Support not confirmed
The evidence-support denominator minus confirmed supported passages. This includes unresolved scoring calls and does not establish that every counted passage is factually wrong or contradicted by its sources.
Passages without citations
An evaluated passage about the subject that gives no source citation. Passages describing only the investigation are excluded.
Unreachable source URLs
A cited URL that returned no page when the grader tried to read it.
Sources marked unread
A cited source marked as unread in the sample report, with only its search snippet available. This count was not recorded for the Deep Research reports.