ChatShield
Sign inGet started
AI chatbot risk monitoring

Your chatbot just made a legal promise you didn't approve.

ChatShield reads every conversation your AI customer-service bot has and flags the exact moments it created liability — a false refund guarantee, regulated advice, exposed personal data, or a missing AI disclosure — before you find out from a lawyer.

Get started See how it works

Works with your existing support stack

Connect in minutes using secure OAuth.

Intercom
Secure OAuth connection
Zendesk
Secure OAuth connection

Built for AI customer support teams — businesses using Intercom and Zendesk AI.

10+
Liability categories detected
100%
Of AI conversations reviewed
OAuth
Secure platform connections
PDF
Exportable audit reports

The Air Canada precedent applies to you.

In 2024, a Canadian tribunal ruled Air Canada was liable for a discount its chatbot invented — even though the correct policy was on its website. The bot's answer was treated as a binding company statement. Your chatbot works the same way.

Contractual promises

The bot invents a discount, warranty, or refund window your policy doesn't allow.

Regulated advice

It answers a medical, legal, tax, or insurance question it isn't licensed to touch.

PII exposure

It repeats a full SSN, card number, or health data back in plaintext.

Missing AI disclosure

It denies being an AI to an EU customer — a violation as of August 2026.

How it works

  1. 01

    Connect Intercom or Zendesk

    Point your Intercom conversations at our webhook via a one-click Make.com scenario.

  2. 02

    AI analyzes every conversation

    Every completed conversation is scored against 10 well-defined risk categories.

  3. 03

    Risk detected instantly

    You get a dashboard of flagged conversations with the exact quote and category.

  4. 04

    Review and export audit logs

    Filter the incident feed and export a PDF audit trail whenever you need one.

What a flag looks like

Sample alerts — illustrative only.

High Risk

Refund promise detected

Medium Risk

Sensitive financial advice

Low Risk

Policy clarification

With and without ChatShield

Without ChatShield
With ChatShield
Refund promises unnoticed
Refund promises detected
Pricing mistakes missed
Pricing mistakes flagged
Regulated advice slips through
Regulated advice categorized
No audit trail
Searchable incident history

Security first.

OAuth authentication

Encrypted transmission

Data minimization

No chatbot modification

Simple pricing

One plan. Cancel anytime.

Most Popular
$99/month

Everything in ChatShield

  • Unlimited Intercom conversations analyzed
  • 10-category liability risk detection
  • Dashboard, filters, and detail views
  • PDF audit log export
  • Email support
  • Secure checkout via Paddle
  • Cancel anytime — no hidden fees
Subscribe

Secure checkout by Paddle. Global tax handled automatically.

Frequently asked

Get started