Make any text safe to paste into AI.
Shield finds passwords, API keys, card numbers, bank details, emails and ID numbers, then swaps them for placeholders like [EMAIL_1]. ChatGPT, Claude and Gemini still understand your text. They just never see the data.
Tip:Add your name or clients under Names to hide them too.
Your text never leaves your device
Shield runs entirely in your browser. Your text is never uploaded, not even to us.
Nothing stored
Your text lives in this tab and is gone when you close it. Only the names you choose to protect are kept, in your browser.
Real numbers only
Card, bank and ID numbers are verified before they are flagged, so ordinary numbers are left alone.
WHAT SHIELD CATCHES
The data that gets people in trouble.
Passwords, API keys, credit cards, bank details, emails and ID numbers. The same engine that runs inside Alesia.
Passwords and keys
Passwords, API keys, access tokens, private keys, database logins
Financial
Credit card numbers, IBANs, bank codes, crypto wallets
ID numbers
Social security, tax and national ID numbers from the US, India, Germany, France, Italy, Spain, the Netherlands, Belgium and Poland
Contact details
Email addresses, IP addresses, and any names you add
CHROME EXTENSION
Shield, right where you type.
Select any text, right-click, and choose Shield. It's protected right there, in ChatGPT, Gmail, Slack or any text box. Same detection, same privacy: nothing leaves your browser.
FROM THE MAKERS OF SHIELD
Shield works when you remember. Alesia works when you don't.
Everything Shield does, Alesia does automatically, for every prompt and upload: in your browser, desktop apps, CLI and IDE. It also shows which AI tools you use and limits what coding agents can do. No copying, no pasting. The checks happen on your device.
FOR YOU
Make every paste safe.
- Free for you and teams of up to 10.
- Catches sensitive data before it reaches ChatGPT, Claude, Gemini, Copilot and other AI tools.
- Works in your browser, desktop apps, CLI and IDE.
- Blocks it before you hit send, or swaps it for a placeholder.
- Runs on your device. Built in Europe.
FOR YOUR TEAM
See and control AI use across the company.
- Every AI tool your team uses, without reading a single conversation.
- Rules per team: which tools, which models, which data.
- Custom keywords and detection models for your own client names and internal IDs.
- Guardrails for coding agents: risky commands and sensitive files stay off limits.
QUESTIONS
Pasting into AI, answered.
What is Shield, and how is it different from Alesia?
Shield is a free tool that hides sensitive data in text before you paste it into an AI tool. You paste, click Protect and copy the result. Alesia does the same automatically, for every prompt and upload, in your browser, desktop apps, CLI and IDE. It also shows which AI tools you use and limits what coding agents can do. Both are free for you and teams of up to 10.
Is it safe to paste personal information into ChatGPT?
Not by default. On free and personal plans, what you type into ChatGPT, Claude or Gemini can be stored, reviewed by people and used to train future models. The AI does not need your passwords, card numbers, bank details or ID numbers to help you, so remove them first.
Source: OpenAI: How your data is used to improve model performance
How do I remove personal data before using ChatGPT?
Use a redaction tool that runs on your device, so the data never reaches a server first. With Shield:
- Paste your text into Shield.
- Check the highlighted values, and add any names you want hidden under Names.
- Click Protect. Each value becomes a placeholder like [EMAIL_1]. Then click Copy.
- Paste the protected text into ChatGPT, Claude or Gemini.
Does Shield upload or store my text?
No. Shield runs entirely in your browser. Your text is never sent to a server, not even ours, and nothing you paste is stored. The page is locked by a security policy that blocks every connection except optional analytics, which never includes your text. Only the names you add are saved, in your own browser.
Will the AI still understand redacted text?
Yes. Shield swaps each value for a labelled placeholder such as [EMAIL_1] or [CREDIT_CARD_1], and the same value always gets the same placeholder. The AI can still follow who and what you mean, and its reply uses the same placeholders.
What sensitive data does Shield detect?
Passwords, API keys, private keys and other secrets; credit card numbers, IBANs, bank codes and crypto wallets; email and IP addresses; and social security, tax and national ID numbers from 9 countries. Card, bank and ID numbers are checksum-verified, so ordinary numbers are left alone.
Does Shield work with Claude, Gemini and Copilot?
Yes. Shield protects your text before you paste it, so it works with any AI tool. The browser extension adds Protect with Shield to the right-click menu, so you can protect text right inside ChatGPT, Claude, Gemini, Copilot or any text box.
Is Shield free?
Yes. Shield is free, with no account and no sign-up. It is made by Alesia, which does the same thing automatically in every AI tool on your laptop, free for individuals and teams of up to 10 people.