Mutant Data Safety Layer

Detect and redact sensitive data before sending to AI tools. Runs entirely in your browser. No uploads. No accounts. No tracking.

Mutant Data Safety Layer scans text for sensitive information before it reaches AI tools like ChatGPT, Microsoft Copilot, Claude or Gemini. Paste an email, document or meeting note. Click Scan. See exactly what's sensitive — and get a clean redacted version ready to paste. Runs 100% locally in your browser. Nothing is uploaded. Nothing is stored. No account required. WHAT IT DETECTS • Email addresses • Phone numbers • National Insurance numbers • Bank account and sort code numbers • Dates of birth • UK postcodes • Named persons • NHS numbers • Passport numbers • IP addresses • Confidential keywords (salary, credentials, NDA, classified...) • Custom protected terms you define SEMANTIC MODE Instead of destroying context, Semantic Mode replaces sensitive data with descriptive labels — so the AI still understands your prompt without seeing the actual data. RIGHT-CLICK TO SCAN Highlight any text on any page. Right-click. Select "Scan with Mutant DSL". Scans instantly. PERMISSIONS Storage only — to save your custom terms. No page monitoring. No telemetry. No data ever leaves your browser. Built by Mutant — mutantwork.com