Guard · Data Protection

Sensitive data held back before a request reaches a model.

Data Protection applies through Guard. Personal and sensitive data is redacted or held back before a request reaches a model, according to the rules you set.

Book the Guard demo Find the right product for you

The problem

Client and personal data gets pasted into AI tools every day, usually without anyone intending harm.

One leak can mean a breach, a regulator, a client conversation and a bill, long before anyone sees the pattern.

£498,000

extra cost on the average UK breach where shadow AI use was high (on a £3.78m average)

IBM, Cost of a Data Breach, UK, 2025 (vendor research)

What you get

The benefits of Data Protection.

Personal and sensitive data redacted before it leaves

Rules by data type, not by hope

A record of what was held back

Who it is designed for

The people who feel the problem first.

DPO and compliance

Reduce the chance of a reportable breach.

Legal

Protect privilege and client material.

CIO and security

A control that works in the flow of work.

Where it is used

Three ways teams use Data Protection.

Customer data

Redact customer details before a request reaches a model.

Client files

Stop confidential documents leaving the business.

HR and payroll

Block sensitive files and notify the owner.

Part of Guard

Works best alongside.

Model Choice

The right model for each task, so premium is by exception.

See Model Choice Rules

Policy as enforced rules: approved models by task, data type and risk.

See Rules Log

Log: the record of events.

See Log All of Guard

Systems, controls and process designed with AI agents and aligned to the regulations, your policies and outside policies. Used with Evidence.

See the Guard module

Prices are indicative ranges, confirmed in writing before you commit. Estimates are replaced by your own figures once your health check runs.

See Data Protection working on your business.