Every day, millions of people share personal information with AI chatbots without thinking twice. PISIGuard is a free, open-source browser extension that automatically spots names, email addresses, API keys, and other sensitive data, replaces them with safe placeholders before they reach the AI, then puts the real values back into the AI's reply. All detection happens locally in your browser. Nothing leaves your device except the placeholders.
PISIGuard sits between you and the AI chat interface. Before your message is sent, it scans for sensitive information and replaces it with safe, context-aware placeholders. When the AI responds, the real values are restored automatically. From your perspective, the process is invisible. From the AI's perspective, the sensitive information was never there.
PISIGuard catches common sensitive information before they reach AI servers: names, email addresses, phone numbers, credit card numbers, API keys, passwords, and more. No manual redaction needed.
Each piece of sensitive data is replaced with a unique, realistic placeholder. For example, your real name becomes a generated name. The AI sees natural-looking text without your actual data.
When the AI responds, PISIGuard automatically puts the real values back into the reply. You read the conversation naturally, with all the correct details restored.
All detection, masking, and restoration happens on your device. Zero data is sent to any external service. No telemetry, no analytics, no server calls. The extension has no background processes.
PISIGuard is fully open source under the AGPL-3.0 license. Anyone can inspect the code, verify the privacy claims, and contribute improvements. No closed-source black boxes.
Advanced users can provide their own set of detection rules for more detailed or custom patterns. PISIGuard adapts to your specific needs.
Install once. Chat normally. PISIGuard works automatically in the background.
Get PISIGuard from the Firefox Add-ons store or load from source on Chrome/Edge. Minimal permissions -- only activates on AI chat sites.
Use ChatGPT, Claude, DeepSeek, or Gemini as you always do. PISIGuard automatically detects and masks sensitive information before your message is sent.
The AI responds to the masked content. PISIGuard restores your real information in the reply. The protection is invisible to you and the AI.
If you use AI chat tools for work or personal matters, PISIGuard helps you maintain privacy without disrupting your workflow.
Paste code snippets with API keys, internal URLs, and database credentials into AI chats without worry. PISIGuard catches and masks them before they leave your browser.
Draft emails, reports, and documents with AI assistance without accidentally exposing client names, contact details, or confidential business information.
Anyone who values their privacy and wants an extra layer of protection when using AI chat platforms. No technical knowledge required.
Learn how PISIGuard compares to alternatives and why privacy matters in AI chats.
Comparison across automation, detection coverage, privacy guarantees, and ease of use.
Read ComparisonUnderstanding what happens to your data when you paste it into ChatGPT and other AI platforms.
Read ArticleWhy local processing matters for true data sovereignty when working with AI tools.
Read ArticleWhy open source is the only trustworthy model for privacy and security tools.
Read ArticleNo. All processing happens locally in your browser. PISIGuard contains no telemetry, no analytics, and no server calls. Your messages are modified before they leave your device and restored when the response comes back.
PISIGuard currently supports ChatGPT, Claude, DeepSeek, and Gemini. The extension activates only on these supported sites and has no background processes otherwise.
Yes. The placeholders are context-aware and realistic. For example, a real name is replaced with a generated name, and an email is replaced with a fake but realistic-looking email. The AI understands the structure of your message without seeing your actual data.
Yes. PISIGuard is completely free and open source under the AGPL-3.0 license. No subscriptions, no paid tiers, no data selling.
PISIGuard is published for Firefox. For Chrome, Edge, and other Chromium browsers, you can install it from the source code on GitHub. Clone the repo, run the build, and load the unpacked extension.
PISIGuard can detect names, email addresses, phone numbers, credit card numbers, passwords, API keys, and other patterns of sensitive information. Advanced users can provide custom detection rules.
No. The AI responds to the masked content naturally. When the response comes back, PISIGuard restores your real information. The protection is invisible to both you and the AI.
VPNs and incognito mode protect your connection and browsing history. They do not protect the content of your messages from the AI platform. PISIGuard specifically protects the sensitive content within your messages.