Watchdog

Tracking since Apr 18, 2026.

Changelog

• May 11, 2026
  description

  Watchdog monitors your conversations with AI assistants and prevents sensitive information from being shared accidentally.
   
  HOW IT WORKS
  When you type a message or upload a file on a supported AI platform, Watchdog scans the content in real time. If it detects personally identifiable information, credentials, or secrets, it gives you three options: receive a warning, automatically redact the data, or block the submission entirely.
   
  WHAT IT DETECTS
  • Email addresses, phone numbers, and dates of birth
  • Credit card numbers, bank account details, and financial identifiers
  • Government-issued numbers (SSN, passport, licence, tax file, Medicare)
  • API keys, access tokens, private keys, and hardcoded credentials
  • Database connection strings, webhook URLs, and bearer tokens
  • GPS coordinates and device metadata embedded in images
   
  FILE PROTECTION
  Watchdog intercepts file uploads before they leave your browser. It scans documents, spreadsheets, presentations, and data files for sensitive content. Detected items can be automatically redacted or blocked. Image files have metadata stripped to remove GPS coordinates and device information. Encrypted files that cannot be scanned are blocked by default.
   
  THREE PROTECTION MODES
  Choose how Watchdog responds when it finds sensitive data:
  — Warn: See exactly what was detected and decide whether to send, redact, or cancel
  — Auto-Redact: Sensitive data is silently masked before your message is sent
  — Block: Messages and files containing sensitive data are prevented from being submitted
   
  PRIVACY BY DESIGN
  On the free tier, all detection runs entirely within your browser. No data is collected, transmitted, or stored on any server. There are no analytics, no tracking, and no network requests. Your data stays on your machine.
   
  FOR ORGANISATIONS
  Watchdog offers paid plans for teams that need centralised visibility. Administrators can deploy the extension across their organisation, set protection policies, monitor detection events, and receive alerts — all from a web dashboard. Telemetry captures only redacted data; original sensitive information is never transmitted.
   
  SUPPORTED PLATFORMS
  ChatGPT, Claude, Gemini, Copilot, DeepSeek, and Perplexity.
   
  Built by CRCYBER — IT Secured by design.
  Learn more at crcyber.com

  Watchdog monitors your conversations with AI assistants and prevents sensitive information from being shared accidentally.
   
  HOW IT WORKS
  When you type a message or upload a file on a supported AI platform, Watchdog scans the content in real time. If it detects personally identifiable information, credentials, or secrets, it gives you three options: receive a warning, automatically redact the data, or block the submission entirely.
   
  WHAT IT DETECTS
  • Email addresses, phone numbers, and dates of birth
  • Credit card numbers, bank account details, and financial identifiers
  • Government-issued numbers (SSN, passport, licence, tax file, Medicare)
  • API keys, access tokens, private keys, and hardcoded credentials
  • Database connection strings, webhook URLs, and bearer tokens
  • GPS coordinates and device metadata embedded in images
   
  FILE PROTECTION
  Watchdog intercepts file uploads before they leave your browser. It scans documents, spreadsheets, presentations, and data files for sensitive content. Detected items can be automatically redacted or blocked. Image files have metadata stripped to remove GPS coordinates and device information. Encrypted files that cannot be scanned are blocked by default.
   
  THREE PROTECTION MODES
  Choose how Watchdog responds when it finds sensitive data:
  — Warn: See exactly what was detected and decide whether to send, redact, or cancel
  — Auto-Redact: Sensitive data is silently masked before your message is sent
  — Block: Messages and files containing sensitive data are prevented from being submitted
   
  PRIVACY BY DESIGN
  On the free tier, all detection runs entirely within your browser. No data is collected, transmitted, or stored on any server. There are no analytics, no tracking, and no network requests. Your data stays on your machine.
   
  FOR ORGANISATIONS
  Watchdog offers paid plans for teams that need centralised visibility. Administrators can deploy the extension across their organisation, set protection policies, monitor detection events, and receive alerts — all from a web dashboard. Telemetry captures only redacted data; original sensitive information is never transmitted.
   
  Built by CRCYBER — IT Secured by design.
  Learn more at crcyber.com

• May 6, 2026
  description

  AI PII Shield automatically detects and blocks personally identifiable information before it reaches major AI chatbots.
  Every day, millions of people paste sensitive data into AI platforms without realising the risk. AI PII Shield acts as a real-time privacy guard between you and your AI assistant — scanning your messages for sensitive information and giving you full control over what gets sent.
  What it detects:
  
  Email addresses
  Credit card numbers
  Phone numbers (formatted and international)
  Social Security Numbers
  IP addresses (IPv4 and IPv6)
  API keys and tokens
  Private keys (RSA, EC, SSH)
  Dates of birth (numeric, written, and context-aware — only flags dates when birth-related context is present)
  Passport numbers, driver's licence numbers, Medicare numbers, and Tax File Numbers
  
  Three protection modes:
  
  Warn — Shows a detailed overlay of every detected item with options to redact, send anyway, or cancel
  Auto-Redact — Silently masks all sensitive data before sending, with no interruption to your workflow
  Block — Prevents any message containing PII from being submitted
  
  Key features:
  
  100% local processing — all detection runs in your browser. No data is ever sent to any server
  Real-time inline badge shows a live count of detected PII as you type
  Built-in test scanner to verify detection without visiting an AI site
  Per-category toggles to enable or disable individual detection rules
  Context-aware detection reduces false positives — dates only flag as dates of birth when relevant keywords are present
  Persistent stats dashboard tracking how many items have been blocked, redacted, or allowed through
  Works across all major AI platforms with automatic input field detection
  Supports InPrivate browsing when enabled in extension settings
  
  Zero tracking. Zero servers. Zero data collection. By design.
  Built by CRCYBER — 24/7 security monitoring for organisations of all sizes.

  Watchdog monitors your conversations with AI assistants and prevents sensitive information from being shared accidentally.
   
  HOW IT WORKS
  When you type a message or upload a file on a supported AI platform, Watchdog scans the content in real time. If it detects personally identifiable information, credentials, or secrets, it gives you three options: receive a warning, automatically redact the data, or block the submission entirely.
   
  WHAT IT DETECTS
  • Email addresses, phone numbers, and dates of birth
  • Credit card numbers, bank account details, and financial identifiers
  • Government-issued numbers (SSN, passport, licence, tax file, Medicare)
  • API keys, access tokens, private keys, and hardcoded credentials
  • Database connection strings, webhook URLs, and bearer tokens
  • GPS coordinates and device metadata embedded in images
   
  FILE PROTECTION
  Watchdog intercepts file uploads before they leave your browser. It scans documents, spreadsheets, presentations, and data files for sensitive content. Detected items can be automatically redacted or blocked. Image files have metadata stripped to remove GPS coordinates and device information. Encrypted files that cannot be scanned are blocked by default.
   
  THREE PROTECTION MODES
  Choose how Watchdog responds when it finds sensitive data:
  — Warn: See exactly what was detected and decide whether to send, redact, or cancel
  — Auto-Redact: Sensitive data is silently masked before your message is sent
  — Block: Messages and files containing sensitive data are prevented from being submitted
   
  PRIVACY BY DESIGN
  On the free tier, all detection runs entirely within your browser. No data is collected, transmitted, or stored on any server. There are no analytics, no tracking, and no network requests. Your data stays on your machine.
   
  FOR ORGANISATIONS
  Watchdog offers paid plans for teams that need centralised visibility. Administrators can deploy the extension across their organisation, set protection policies, monitor detection events, and receive alerts — all from a web dashboard. Telemetry captures only redacted data; original sensitive information is never transmitted.
   
  SUPPORTED PLATFORMS
  ChatGPT, Claude, Gemini, Copilot, DeepSeek, and Perplexity.
   
  Built by CRCYBER — IT Secured by design.
  Learn more at crcyber.com

• May 6, 2026
  short_description

  By CRCYBER — Detect and block personally identifiable information before it reaches AI chatbots. 100% local processing.

  By CRCYBER — Detect and block sensitive data before it reaches AI chatbots. Your data watchdog. 100% local processing.

• May 6, 2026
  name

  AI PII Shield

  Watchdog

• May 6, 2026
  category

  productivity/tools

  make_chrome_yours/privacy

• May 6, 2026
  permissions

  storage, activeTab

  storage, activeTab, alarms

Permissions & access

Permissions

storageactiveTabalarms

Host access

None declared

Screenshots