Security
An AI agent you can trust with customer data and real actions.
An agent that books and changes things needs more than good answers. Loopwix verifies customers first, does only what you permit, and records everything it does.
Beta opens end of October 2026. Early access is invite-only.
In short
Loopwix shares personal data only after the customer is verified, every capability is a permission you switch on or off, high-risk actions can require confirmation or staff approval, and every action is written to an audit trail.
Want the thinking behind these controls? Read about Loopwix.
Last updated 10 October 2026
Identity
Customers are verified before anything personal is shown.
Order history, bookings and billing details only appear after the customer proves who they are, with an email code, a magic link or a signed sign-in from your own site.
Permissions
Every capability is a switch you control.
Reading orders, booking and cancelling are separate permissions. Higher-risk actions can ask the customer to confirm, or wait for your team to approve, before they happen.
Audit trail
A permanent record of what the agent did.
Every lookup and action is written to an audit trail, so you can review what happened and when, and your team can pick up any conversation with the full context.
Under the hood
Six protections that work together.
You decide how strict each one is.
Verify before personal data
Email code, magic link or signed sign-in before anything personal is shown.
Permissions are switches
If you switch a capability off, the agent cannot do it.
Confirmation and approval
Higher-risk actions can need customer confirmation or staff approval.
A permanent audit trail
Every action is logged so you can review what happened and when.
Workspace isolation
Your data stays in your workspace and is not used to train shared models.
Scoped system access
Credentials are encrypted and the agent only calls what you allow.
Data handling
- Retention. Conversation history is kept for a configurable period (default 90 days) and deleted on schedule.
- Privacy. How we handle information on loopwix.com and in the platform is described in our Privacy Policy.
What is on the roadmap
Not available yet
- Regional hosting for EU customers
- Single sign-on and extended compliance tooling
- A published Data Processing Agreement and subprocessor list (arriving as early access rolls out)
We would rather tell you what is not ready than imply it is. If you need something specific before you can adopt Loopwix, or you have a security questionnaire, email [email protected].
Security questions
When does Loopwix share personal data with a customer?
Only after the customer is verified. Verification is by an email code, a magic link or a signed sign-in from your own site, and it happens before order history, billing details or any other personal data is shown.
Can I stop the agent from doing certain things?
Yes. Every capability is a permission you switch on or off, such as reading orders or cancelling them. Higher-risk actions can require the customer to confirm, or require staff approval.
Is my data used to train shared AI models?
No. Each workspace is isolated: your knowledge, connection credentials and conversations stay scoped to your account and are not used to train shared models.
How long are conversations kept?
Conversation history is kept for a configurable period, with a default of 90 days, and deleted on schedule to support data-minimisation requirements.
Does Loopwix have a Data Processing Agreement and a subprocessor list?
A Data Processing Agreement and a subprocessor list will be published as early access rolls out. Until then, email us for security or vendor questionnaires.
Join the waitlist.
Early access is invite-only and includes all features. Tell us where to send it.