Policy Change Detector — Privacy & Terms Monitoring

Monitor privacy policies, terms of service, and legal pages for meaningful changes. Extract policy text, detect policy type, compare versions, classify changes by category, and assess severity.

The Problem

Legal pages change. When a company updates their Privacy Policy or Terms of Service, the changes can be significant — new data sharing, different retention periods, removed user rights. But detecting meaningful changes in legal pages is hard because:

• Raw HTML changes constantly (CSS, navigation, scripts)
• Generic change detectors just say "something changed"
• You need to know WHAT changed and HOW IMPORTANT it is

This Actor tells you: "This policy changed, here are the changed sections, these changes may relate to data sharing and retention, review may be recommended."

What It Does

Policy URL
    ↓
Fetch HTML
    ↓
Extract policy text (remove scripts, styles, nav, footer, cookie banners)
    ↓
Detect policy type (privacy_policy, terms_of_service, etc.)
    ↓
Normalize text (collapse whitespace, normalize Unicode)
    ↓
Generate content hash (SHA-256)
    ↓
Compare with previous version (if provided)
    ↓
Detect added/removed/modified sections (difflib)
    ↓
Classify changes by category (13 policy categories)
    ↓
Assess severity (low/medium/high)
    ↓
Push structured results to Apify Dataset

Two Operating Modes

Mode 1 — Analyze Current Policy

Fetch a policy page, extract text, detect type, generate hash.

{
  "urls": ["https://example.com/privacy"],
  "compareWithPrevious": false
}

Mode 2 — Compare Policies

Compare previous content with the current version.

{
  "urls": ["https://example.com/privacy"],
  "previousContent": "Previous policy text here...",
  "analyzeChanges": true
}

Or use oldUrl + newUrl for explicit comparison:

{
  "oldUrl": "https://example.com/privacy-old",
  "newUrl": "https://example.com/privacy"
}

Output Example

{
  "url": "https://example.com/privacy",
  "success": true,
  "policyType": "privacy_policy",
  "title": "Privacy Policy",
  "contentHash": "abc123...",
  "changed": true,
  "changeSummary": "Potentially significant changes detected (2 changes). Review recommended.",
  "severity": "high",
  "changes": [
    {
      "category": "data_sharing",
      "changeType": "added",
      "severity": "high",
      "summary": "New Data Sharing language was added.",
      "oldText": null,
      "newText": "We may share information with advertising partners."
    },
    {
      "category": "data_retention",
      "changeType": "modified",
      "severity": "medium",
      "summary": "Data Retention section was modified.",
      "oldText": "We retain information for 30 days.",
      "newText": "We retain information for 12 months."
    }
  ],
  "checkedAt": "2026-09-02T00:00:00Z"
}

Change Categories (13)

Category	What It Detects
data_collection	What data is collected
data_sharing	Sharing with third parties
third_party_services	Service providers, vendors
analytics_tracking	Analytics, tracking, pixels
data_retention	How long data is stored
cookies	Cookie policy, local storage
user_rights	Access, deletion, opt-out rights
security	Security measures, breach notification
international_data_transfer	Cross-border data transfers
children_data	Data about children/minors
marketing	Marketing, promotional emails
legal_basis	Legal basis for processing
other	Uncategorized changes

Severity Levels

Level	Examples
high	New third-party data sharing, new sensitive data collection, user rights removed, international transfer changes
medium	Retention period changed, cookie policy changed, analytics section changed
low	Minor wording changes, formatting, typo corrections

The Actor does NOT claim legal violations. It uses language like "potentially significant" and "review recommended."

Use Cases

• Vendor monitoring — Track vendor privacy policy changes
• Compliance research — Monitor regulatory compliance across policies
• Privacy research — Track data sharing and retention changes
• Competitor monitoring — Watch competitor policy changes

Input

{
  "urls": ["https://example.com/privacy", "https://example.com/terms"],
  "previousContent": "Previous policy text...",
  "detectPolicyType": true,
  "extractText": true,
  "compareWithPrevious": true,
  "analyzeChanges": true,
  "maxUrls": 100,
  "requestDelayMs": 500
}

Pricing

Pay-per-event: $0.05 per policy-analysis event

Charged for each successfully analyzed URL.

Limitations

• The Actor does not guarantee legal compliance
• The Actor does not claim legal violations
• AI analysis is optional and not required for basic functionality
• Structured-data interpretation may change over time
• The Actor detects changes; it does not determine whether a change was intentional

License

MIT