How BeeSensible works

BeeSensible highlights sensitive data while you type in AI tools, email, and chat, in Chrome and Edge. You decide what happens to it. The extension never changes your text on its own and never blocks sending.

#Highlights while you type

When you type or paste something with a sensitive value in it, a coloured underline appears under that word:

• Yellow is a warning, for ordinary data such as a name or email address.
• Red is for highly sensitive data, such as a credit card number or IBAN.

The highlight sits as a layer on top of your text; your draft stays exactly as you wrote it. The panel on the page keeps count and shows how many sensitive items are in your message.

#Three actions per highlight

Click a highlight and choose:

• Replace: the real value becomes a made-up but realistic-looking one. The structure of your text keeps making sense.
• Mask: the value becomes a label such as [NAME] or [IBAN]. The kind of data stays visible, the content does not.
• Delete: the value comes out of your text.

You can also apply an action to everything at once. See Review detections and replace, mask, or delete for the full walkthrough.

#Sending stays your call

BeeSensible never blocks sending. If sharing something is fine in this context, you send your message as usual. The reverse holds too: detection can miss something, so review your message yourself before you send it.

#Where detection runs

Your organisation decides this, and you can see it at the bottom of the extension settings:

• Detection: on-device: detection runs through the BeeSensible desktop app on your own computer. The text you type does not leave your device.
• Detection: cloud: the text goes to BeeSensible's servers in the EU, is checked there in working memory, and is discarded right after. Nothing you type is stored.

In both cases, detection stays within BeeSensible. No external AI services such as OpenAI, Anthropic, or Google are called.

#Notices about AI tools

If you open an AI tool your organisation does not allow, or has not decided on yet, a notice can appear. It tells you what is going on and points to an approved alternative when one is set. This notice does not block anything either. See When BeeSensible flags an AI tool.

#Your settings

Open the panel and click the settings icon. Two settings are yours:

• Language: System, English, or Nederlands.
• Theme: System, Light, or Dark.

Below that you find your organisation, the detection mode, and the extension version. On a managed workplace, a Managed by your organisation card takes the place of the logout button: signing out or switching accounts goes through your IT department.

#Supported apps