Redactprompt

Prevent data leaks to AI chatbots.

Redactprompt catches personal data and API keys in your prompts before they reach AI chatbots likes ChatGPT, Copilot, Gemini, or Claude. It all runs in your browser, so nothing sensitive ever leaves your device.

Everyone pastes things into AI tools they probably shouldn't: a customer's BSN, a support ticket with someone's address, an API key copied straight out of a config file. Redactprompt watches for that as you type or paste, and catches it at the last moment, before you hit send.

HOW IT WORKS

The moment your prompt contains something sensitive, Redactprompt shows you exactly what it found and where. Nothing has been sent yet. You choose what happens next:

• Redact: replace the sensitive parts with clear labels like [BSN REDACTED], so the AI still understands the shape of your question
• Pseudonymize: swap real data for realistic but fictional stand-ins, so the model has enough context to still give a useful answer
• Send as-is: if it's a false positive, keep your original text and carry on

Your choice is remembered, so once you've set a preference it stays out of your way.

WHAT IT CATCHES

Personal data: BSN, IBAN, email addresses, phone numbers, postal codes, dates of birth, passport and ID numbers, and more, tuned for both Dutch and English text.

Secrets and credentials: API keys, tokens and passwords from OpenAI, AWS, GitHub, Stripe, Google, and 60+ other providers, matched by pattern and entropy so real keys get flagged and random strings don't.

All of it runs on-device in a few milliseconds, while you type.

PRIVACY FIRST

Everything that matters happens on your device. Detection runs in your browser, and no prompt text is ever sent to our servers, on any plan, for anyone. When an organization uses the premium or enterprise dashboard, it receives usage metadata only: how many detections, in which categories, on which chatbots. Never the words you typed.

Because your prompts never leave the browser, Redactprompt also helps you stay on the right side of GDPR, the EU AI Act, NIS2, and DORA.

START FREE, UPGRADE FOR YOUR ORGANIZATION

Install Redactprompt and use it on your own, free, with no account and no credit card. All the detection works exactly the way it does for paying customers. When you need to protect a whole team instead of just yourself, that's where the organization plans come in.

BUILT FOR ORGANIZATIONS

Roll Redactprompt out across your company and manage it from a single dashboard. You can:

• Set detection policy centrally for everyone, or let people adjust it within limits you define
• Block chatbots you don't allow, and switch off risky features like file upload or chat sharing per chatbot
• Add your own detection rules for internal identifiers, project codenames, or anything specific to your business
• See where and how AI is being used across the organization, without ever collecting the prompts themselves
• Deploy it silently through Intune, with SSO and SAML

WORKS WITH

ChatGPT, Microsoft Copilot, Gemini, Claude, and a growing list of other AI chatbots are supported. The extension also works with AI-powered search engines such as Google and Bing. It is compatible with all Chromium-based browsers, including Chrome and Edge.

AT A GLANCE

• Real-time detection while you type or paste
• Redact, pseudonymize, or send as-is, your call every time
• 17 personal-data detectors, Dutch and English
• 85 secret and credential patterns across 60+ providers
• 100% local processing, no prompt content sent anywhere
• Free for individuals, no account needed
• Central dashboard, policies, and per-chatbot controls for teams
• Custom detection rules
• SSO / SAML and Intune rollout
• Made in Europe

Get started in under a minute.