Policy Radar

Policy changes

Review detected policy changes in a sortable table, filter by severity and vendor, view redline diffs, and create risks.

Overview

The Changes page collects every policy change found across your monitored vendors. Each time a scan picks up that a page has changed, Policy Radar creates a change record with an AI-generated analysis of what is different. The page title is "Changes" and the URL path is /policy-radar/events.

Changes table

Changes are displayed in a sortable table. Click any column header to sort ascending or descending. The table columns are:

  • Vendor: The vendor name with its logo. Sortable alphabetically.
  • Page: Which policy page changed (e.g., Terms of Service). Sortable alphabetically.
  • Type: The document type shown as a chip (Terms of Service, Privacy Policy, etc.). Sortable.
  • Severity: AI-classified level shown as a colored chip: critical, high, medium, or low. Sortable by severity rank.
  • Summary: A short AI-generated description of the key changes. Not sortable.
  • Detected: When the change was first detected. Sortable by date, defaults to newest first.
  • Status: Whether the change is new, reviewed, acknowledged, dismissed, or auto-dismissed. Sortable.

Click any row to open its detail page for the full analysis and diff.

Filtering and searching

Three controls at the top of the page help you find specific changes:

  • Search: Type in the search box to filter by vendor name, page name, or summary text
  • Severity filter: Choose a severity level or a threshold: "Medium and above", "High and above", "Critical only", individual levels, or "All severities". The default matches your alert threshold setting.
  • Vendor filter: Show changes from one specific vendor, or all vendors
Quick triage
Set the severity filter to "Critical only" to jump straight to the changes that need immediate attention. Auto-dismissed low-severity changes are hidden unless you select "All severities" or "Low only".

Severity levels

The AI classifies each change into one of four severity levels:

  • Critical: Changes to data handling, liability, or indemnification that could directly affect your compliance posture
  • High: SLA, data processing, or privacy practice changes that warrant prompt review
  • Medium: Scope or definition updates to review in your next cycle
  • Low: Wording or formatting tweaks with no material impact. These are auto-dismissed by default.

Reviewing a change

Click any row to open the detail page. At the top you will find an analysis overview card with:

  • Severity: The classified severity level as a colored chip
  • Change magnitude: A visual progress bar showing how significant the change is as a percentage, color-coded from green (small) through amber (moderate) to red (large)
  • Confidence: How confident the AI is in its classification, shown as a percentage
  • Frameworks: Compliance frameworks that may be affected, shown as chips (e.g., EU AI Act, ISO 42001, GDPR)

Below the metrics you will find the AI summary describing what changed and why it matters, along with a recommended action when applicable.

Diff viewer

Two viewing modes are available, toggled via tabs:

  • Redline: A legal redline view that shows sentence-level changes inline. Added text is highlighted in green and removed text in red, similar to tracked changes in a word processor.
  • Side-by-side: A split view using react-diff-viewer showing the old text on the left and new text on the right, with word-level highlighting. A "Show full diff" link expands the view if the text is long.

Version history

Below the diff you will find the version history for this target. It shows all previously captured snapshots of the same page, so you can track how the policy has evolved over time.

Taking action on a change

An Actions dropdown in the top right of the detail page provides four options:

  • Mark as reviewed: Signals that someone has reviewed this change and no further action is needed
  • Acknowledge: Indicates the change has been seen and noted, but may still need follow-up
  • Dismiss: Marks the change as not relevant to your organization
  • Create risk: Creates a new risk in your risk register, pre-filled with the vendor name, change summary, and severity

The current status is indicated next to its label in the dropdown. When a risk is successfully created, the button updates to "Risk created".

If AI analysis was unavailable during the scan (no API key configured), a warning banner appears at the top of the detail page. The diff is still available, but severity, summary, and recommendations will not be generated.
PreviousManaging monitored vendors
NextPolicy Radar settings
Policy changes - Policy Radar - VerifyWise User Guide