Skip to content
lussy

A product by Lussy

Use AI. Keep the data.

CloakAI masks personal data and secrets before a prompt is sent, then restores them in the reply.

Prompt to your AI assistant

Summarise this complaint from Priya Sharma[PERSON_1] (priya.sharma@northwindmail.co.uk[EMAIL_1], account 40318827[ACCOUNT_1]). She says the refund to her card ending 4417[CARD_1] never arrived and asks us to call 07700 900 418[PHONE_1].

5 sensitive values detected. Sent as written.

Try the switch. This is a working example of what CloakAI sends to the model.

How it works

  1. Detect

    Names, emails, phone numbers, addresses, account and card numbers, API keys and your own custom patterns are found as you type.

  2. Mask

    Each value is swapped for a consistent placeholder, so the model can still reason about who said what without ever seeing it.

  3. Restore

    Real values are put back into the reply on your device. Your team reads a normal answer.

Built by a security team, for teams that handle customer data

Common questions

Which AI tools does CloakAI work with?

CloakAI is designed to work with mainstream AI assistants and APIs. Request a demo and we will confirm support for the tools your team uses.

Does CloakAI see our data?

Detection and masking are designed to run before a prompt is sent, so the sensitive values are replaced before they reach any third party, including us.

Does masking make the AI less useful?

Rarely. Placeholders are consistent within a conversation, so the model can still follow who is who and what refers to what. It just never learns the real values.

Can we try it?

Yes. Request a demo and we will walk you through it with examples from your own workflows.

See CloakAI with your own workflows.

A 30-minute walkthrough using examples from your team. No slides, no commitment.