AI Shield — Prompt & Data Guard

About

AI Shield — Prompt & Data Guard
 
AI Shield runs silently in your browser and alerts you before sensitive information gets pasted into AI chatbots or messaging platforms. Every scan happens locally on your device — nothing is ever sent to any server.
 
THE PROBLEM
 
When you copy-paste text into AI assistants, you may unknowingly share credit card numbers, passwords, API keys, or other sensitive data. Worse, the text you paste could contain hidden prompt injection attacks designed to manipulate the AI. AI Shield catches both risks before they reach the chat.
 
WHAT IT DETECTS
 
Sensitive data (PII):
Credit card numbers with Luhn validation, Social Security numbers, email addresses, phone numbers, IBAN bank account numbers, API keys for popular developer platforms, passwords, and authentication tokens.
 
Prompt injection attacks:
Over 30 known attack patterns in English and French, including role hijacking, instruction overrides, and encoded payloads.
 
Hidden content:
When you paste rich text, AI Shield inspects the underlying HTML for invisible content — zero-size fonts, hidden elements, and transparent text that could carry malicious instructions you cannot see.
 
HOW IT WORKS
 
1. Install AI Shield and browse normally.
2. When you type, paste, or drag content into a supported site, the extension silently scans it in milliseconds.
3. If a threat is detected, AI Shield blocks the submission and shows a clear alert explaining exactly what was found.
4. You decide: edit your message, dismiss the alert, or cancel. You stay in control.
 
KEY FEATURES
 
Real-time scanning of typed, pasted, and dropped content.
Works on popular AI assistants and web-based messaging platforms.
Blocks submission when threats are detected, with clear and actionable alerts.
File scanning before upload — drag and drop any text file to check it.
100% local processing — zero data leaves your device.
No account required. No analytics, no tracking, no cookies.
Dark mode support with a clean, minimal interface.
Available in English and French.
Built on Chrome Manifest V3 for performance and security.
 
PRIVACY
 
AI Shield is built with a strict privacy-first approach:
 
All analysis runs inside your browser — nothing is transmitted to any server.
No personal data is collected, stored, or shared.
No external API calls for scanning.
No user accounts or sign-ups required.
Minimal browser permissions — only what is strictly needed to scan input fields.
 
Full privacy policy: https://aidtaskpro.com/ai-shield/privacy/
 
AI Shield is completely free. Your privacy matters.
 
SUPPORT
 
Questions or feedback? Contact us at [email protected]
Learn more: https://aidtaskpro.com/ai-shield/
 
Built by AidTaskPro — practical tools for digital security and productivity.