Data Guard автор Bhuvaneshwari_Data_Guard

Data Guard is a privacy-focused browser extension that helps you avoid accidentally oversharing sensitive personal information with AI chat tools.

As AI chatbots become part of daily life, people often paste or type sensitive details — social security numbers, credit card numbers, passwords, home addresses, and health information — without realizing that data may be stored, logged, or used for training by the AI provider. Data Guard watches your input in real time on supported AI chat sites and shows a clear warning before you send a message containing potentially sensitive data.

WHAT IT DETECTS
- Social Security Numbers
- Credit card numbers
- API keys
- Phone numbers
- Street addresses
- Passwords
- Health-related information

HOW IT WORKS
Data Guard runs entirely inside your browser. It checks the text in the active chat input box against a set of detection patterns and displays a non-blocking warning banner if something risky is found. You always remain in control — the extension never blocks or modifies your message, it simply informs you before you decide to send it.

PRIVACY BY DESIGN
Data Guard collects zero data. Nothing you type is ever sent to a server, logged, or stored — locally or remotely. All detection happens using local pattern matching directly in your browser.

SUPPORTED SITES
- ChatGPT (chat.openai.com, chatgpt.com)
- Claude (claude.ai)
- Gemini (gemini.google.com)

This is an early-stage, open-source project built to explore practical, user-facing safeguards against AI-related privacy risks. Detection is pattern-based and may not catch every case — see the project's GitHub repository for an honest evaluation of its accuracy and known limitations.

Source code and evaluation results: https://github.com/727724eucd010-Bhuvaneshwari/Data_guard