Vero logo
Local privacy Vero
Prompt shielding

Protect prompts before the model ever sees raw data.

Vero scans messages inside supported AI interfaces, replaces sensitive values locally in the browser, and keeps the interaction deliberate instead of noisy.

Public install link not live yet. Use the support page to request access or publication updates.

Processing
Local PII redaction runs inside the browser extension, not on a Vero server.
Scope
Prompt-first Email, phone, IBAN, VAT, national numbers, API keys, money, and named entities.
Design
Quiet Protection is automatic. The extension stays out of the way unless you ask otherwise.
Trust model

Privacy posture, stated plainly.

Full policy

No Vero cloud relay

Vero does not proxy your prompts through its own backend. Detection runs inside the extension before your browser submits the anonymized prompt to the AI website you chose.

Temporary local mapping

Placeholder mappings are stored locally in extension storage to support protected output rendering, then cleaned up on tab close or after a limited retention window.

Layered detection

Regex rules cover structured values quickly. Local NER analysis adds semantic detection when the bundled model is available.

User control

Protection can be enabled or disabled from the extension UI. Discreet display mode lets you blur restored values until hover.

Audience

For people who use AI tools in environments where leakage matters.

Vero fits teams and individuals who work with logistics, operations, finance, support, legal, HR, customer data, or codebases with secrets that should never leave the browser in raw form.

Positioning

Protection should feel intentional, not theatrical.

The product language and interface stay compact on purpose. Vero should add friction only where privacy risk exists, and disappear when it does not.