Skip to main content

Global Sensitivity Control for Compliance Scans

Control how strictly Warrant’s AI flags potential compliance issues during reviews and scans.

Written by Austin Carroll

Review Sensitivity controls how strictly Warrant's AI flags potential compliance issues during reviews and scans. Admins can adjust this setting to match their organization's risk tolerance, from minimal flagging to near-exhaustive detection.

How It Works

Use the slider to select one of five preset sensitivity levels. Higher sensitivity shows more lower-confidence findings. Lower sensitivity shows only the findings Warrant is most confident about.

The system updates immediately; no manual save is required. As you move the slider, a guidance panel explains what that level will surface. New workspaces start at Mild.

The Relevancy Confidence Auto-Dismissal switch on the same page follows this setting. When it is on, findings that fall below your chosen sensitivity level are dismissed automatically. Turn it off to require human review of every finding.

Findings from a policy your reviewers have repeatedly marked not relevant can still be dismissed automatically, even with this switch off. See Working with Scan Findings.

Where to Find It

  1. Click Settings in the left sidebar.

  2. Under Compliance, select Review settings.

  3. In the AI compliance section, find Review sensitivity. It appears when Enable AI compliance scanning is on.

Sensitivity Levels

  1. Flexible — Flags only the most serious, high-confidence issues. Best when speed matters and content is already well controlled.

  2. Mild — Flags high and medium-confidence issues. Recommended starting point for most organizations.

  3. Balanced — Catches most potential issues with a balance of precision and coverage across compliance areas.

  4. High — Surfaces possible concerns that may need human review. Recommended for highly regulated industries.

  5. Maximum — Flags nearly everything for review, including minor or low-confidence findings. Best when risk tolerance is very low.

Where It Applies

This setting affects AI reviews across:

  • Studio

  • Scan

  • Automated issue detection across assets

All future reviews use the selected sensitivity level.

Note: Brand Guideline checks (style, tone, brand rules) now have their own independent sensitivity, set separately with Brand guidelines sensitivity on the same Review settings page. It starts at Same as compliance sensitivity (default), which follows this slider. Once you choose a different level there, this slider controls only compliance/regulatory flagging.

Did this answer your question?