Privacy audit report

The Privacy audit report (Privacy Report) shows how sensitive data is protected. It works as a pre-check and as an after-the-fact audit snapshot. It lists privacy status per column and includes a Privacy rank for each column in the exported CSV.

When to use the Privacy Report

Before running a job (pre-check)

Export a Privacy Report to confirm that:

• Sensitive columns are correctly identified
• Generators are applied where needed
• Consistent mapping matches your risk posture
• No sensitive columns remain Unprotected

After a job completes (verification)

A Privacy Report is generated automatically after each job so you can confirm:

• The privacy status at the time the job ran
• The Privacy rank captured for each column when the job ran
• The applied generator configuration used for the job
• Any remaining sensitive columns that may require follow-up

What’s included in the report

The Privacy Report is exported as a CSV file containing one tab with a row for every column in the workspace.

Each row includes:

• Workspace mode
• Schema
• Table
• Column
• Data type
• Sensitive (True/False)
• Sensitivity scan suggestion (only shown if a suggestion exists and was not rejected)
• Applied generator
• Consistent mapping enabled (True/False)
• Column privacy status, one of:

  • Unprotected
  • Protected
  • Non-sensitive
  • Excluded table
• Privacy rank (integer 1–6, where lower values mean stronger protection)

Privacy rank is the column-level privacy score in the report. Excluded tables may have an empty Privacy rank because no generated output is produced for those columns. For details, see Privacy ranks.

Exporting the report

Export anytime (current workspace state)

This export reflects the protection level at the moment you export it.

1. Open the workspace you want to review.
2. Go to the Main hub in the job configuration.
3. In the Main hub panel, export the Privacy Report.

Export for a specific job (job-time snapshot)

A Privacy Report is automatically created after each data generation job and reflects the protection level at the moment the job ran.

You can export it from:

• The Job summary panel
• The Jobs panel, after selecting one or more job(s)

You can export the Privacy Report without generating data. Use it as a configuration and privacy check whenever needed.

Interpreting “Column privacy status”

Use this as a quick triage label. The report is still column-level, not “dataset certified”.

• Unprotected: The column needs protection, but none is applied.

  • Typical causes: Sensitive=True and no generator. Or a risky generator choice.
• Protected: A protection method is applied.

  • Example: Mock, Mask, AI synthesize, or another configured generator.
• Non-sensitive: The column is not marked as sensitive.
• Excluded table: The table is excluded from the job configuration.

What to do next

1. Filter the report to Column privacy status = Unprotected.
2. Fix the column configuration.

   • Re-run the sensitivity scanner if sensitivity detection looks off. You can also re-assign sensitivity statuses manually.
   • Apply a generator or exclude the column/table.
3. Export again to confirm the status changed.

Known considerations

The report depends on correct sensitivity identification and configuration. If either is wrong, statuses can mislead. Review Unprotected columns before running jobs.