Is this text safe to paste into ChatGPT?

Is my data safe in ChatGPT? Check before you send: paste your text here. The free test instantly shows which sensitive data it contains, highlights the passages and explains for each data type why it does not belong in an AI chat. No sign-up, no installation.

How the test works

1. Paste text: for example a customer email, meeting notes or a piece of code you want ChatGPT to rework.
2. Look at the highlights: every match is highlighted in the text. Below, each data type shows how often it occurs and why it is sensitive. Passages with a dashed underline are possible matches, such as a number without a matching keyword.
3. Replace before sending: replace the passages with placeholders such as [IBAN] or [Customer]. The AI almost never needs the real values to answer.

What is detected?

The test uses the same rules as the extension: bank data (IBANs, card numbers), social security and tax numbers from Switzerland and Europe, identity documents, contact details and credentials such as API keys, tokens and passwords in configurations. Numbers with a check digit are recalculated so that random numbers do not trigger. The full list with official test values is in the documentation under What is detected.

This test does not detect names, addresses, health details in running text or trade secrets. The extension additionally detects names, companies, postal addresses and health details with a local AI model. Still, read your text yourself as well.

Is my text sent or stored?

No. The check runs as a script in your browser. The page has no form, sets no cookies and does not store the text anywhere. Its Content Security Policy does not allow a single connection (connect-src 'none'), so the browser would block any request. Spell checking and automatic translation are switched off for the text field.

Check it yourself: open the developer tools (F12), switch to the “Network” tab and paste a text. No request appears while the check runs.

What the page cannot prevent: installed browser extensions, such as grammar or translation helpers, can read text in input fields. If you use such extensions, check sensitive texts in a private window without extensions.

Is it safe to put customer data in ChatGPT?

Usually not without precautions. Typing personal data into an AI service passes it on to a provider, often abroad. The company needs a legal basis and a suitable contract with the provider for that; private accounts have neither. Read more in the guides Customer data in ChatGPT and IBAN and ID numbers in AI tools. For clear rules in your team there is a free AI policy template.

Frequently asked questions

Is my data safe in ChatGPT?

It depends on the data and the account. Bank details, ID numbers, passwords and customer data do not belong in a chat with a private account. Check your text here first and replace such passages with placeholders.

Is the test free?

Yes. The test is free, with no sign-up and no email address. The AImori extension is free as well.

Does the test also work for Copilot, Gemini and Claude?

Yes. The test checks the text itself, not the AI service. Whatever is highlighted here would be sensitive in any AI chat.

Does “No known patterns found” mean the text is safe?

No. The test detects numbers, keys and contact details with a fixed format, but not names, diagnoses or confidential content in sentences. Names and health details are only detected by the extension with its local AI model.

Why install the extension if there is a test?

The test helps when you remember to use it. The extension checks every prompt directly in the AI chat before it is sent, even when you are in a hurry.