AI Prompt Anonymizer

The AI Prompt Anonymizer is a free, browser-based tool that detects and masks personally identifiable information (PII) and secrets in text before you share it with AI assistants like ChatGPT, Claude, or Gemini. It identifies 15 categories of sensitive data, including emails, phone numbers, API keys, and credit card numbers, and offers three masking modes: reversible tokens, solid redaction, or realistic fake data. Designed for developers, support staff, and privacy-conscious users, it runs entirely on your device, ensuring nothing is uploaded or logged.

How it works

Paste your prompt or document into the tool, and it instantly highlights detected sensitive data with a live risk score and per-category breakdown. You can toggle categories on or off and add custom terms to always mask. Choose between token mode, redaction bars, or fake data, then copy the clean version to paste into any AI assistant. If you use token mode, you can later paste the AI's reply into the Restore tab to swap placeholders back to your real values.

Key capabilities

The tool detects 15 categories including email addresses, API keys and tokens, credit card numbers (Luhn-validated), US Social Security numbers, IBANs (mod-97 validated), crypto wallets, IP and MAC addresses, phone numbers, URLs, street addresses, social handles, person names, and dates. Person names and dates are off by default because they are the noisiest detectors. It also supports custom always-mask terms for values not covered by built-in patterns.

Who it is for

It is ideal for redacting customer names, emails, and order numbers from support tickets before drafting replies with AI, stripping API keys and IP addresses from error logs and stack traces, anonymizing contracts, medical notes, or HR documents, and removing personal data from spreadsheets or CSV snippets. It also supports GDPR data minimization and HIPAA de-identification workflows as a first step, though it is not a compliance guarantee.

Pricing and limits

The tool is free with no usage limits and no signup required. All detection and masking happen locally in your browser, so it works offline once loaded and leaves no trace when you close the tab. Detection is pattern-based rather than an AI model, making results predictable and repeatable, but it only finds what it has a pattern for. The token map lives only in the browser tab's memory and is never saved, so restore replies before closing or reloading the page.