If your organisation has enabled AI analysis, a case can be analysed with one of the analysis prompts your admin has configured — for example a triage summary or a risk assessment aligned to your local whistleblowing legislation.

​
Running an analysis

  1. Open the case and choose an analysis prompt.
  2. The report text is read live from the intake system and analysed.
  3. The result is saved to the case as an internal note, and the run itself is recorded in the audit trail.
The output is a suggestion for the analyst — it never changes the case by itself, and it is excluded from exported reports like any other internal note.

​
Where the analysis runs

Your admin chooses the engine in Settings → AI & Prompts:
  • In your browser — the model runs entirely on your device; report text never leaves it. The first run downloads the model, so it can take a little while.
  • Server-side — an organisation-configured AI endpoint processes the text; the connection is managed centrally and no keys are exposed to your browser.
Which prompts you can run depends on your role’s access, and prompts can be linked to the intake questionnaire your organisation uses — so you only see prompts written for the form the reporter actually filled in.