Documentation Index

Fetch the complete documentation index at: https://knowledge.catonetworks.com/llms.txt

Use this file to discover all available pages before exploring further.

ChatGPT Integration with AI Security

  • Updated on Jul 6, 2026
  • Updated on Jun 2, 2026
  • 1 minute(s) read
Prev Next

Create a ChatGPT connector in the Cato Management Application (CMA) to monitor prompts with AI Security.

Overview

Cato AI Security uses the ChatGPT Enterprise Compliance API to monitor prompts.

  • Unified AI Auditing - AI Security uses the ChatGPT Enterprise Compliance API to consume conversation history and add it to the unified AI engine. Our customers use this auditing for security detections, adoption analytics, compliance, and eDiscovery.
  • Prompt detection and response - AI Security runs a detection engine with pre-defined signals for risky interactions and customizable data detections.
  • AI asset management - AI Security creates a list of all GPTs and other types of AI assistants.
  • Knowledge and capabilities mapping and risk management - AI Security graphs the relationships between users, assistants, knowledge, and the outer world to clarify the risks.
  • Scanning for GPTs misconfigurations and remediation - AI Security scans your AI assets with a predefined list of over 100 types of misconfigurations found by

Creating a ChatGPT API Key for AI Security

To enable prompt monitoring for the OpenAI connector, create a new API key for the administered OpenAI organization and request the required Compliance API scopes from OpenAI. Cato AI Security uses this access to retrieve prompt data from the workspace for inspection and policy enforcement. For more information, see the OpenAI documentation.
Best Practice: Create a Service Account API key where available rather than a user-owned API key.
To create a ChatGPT API key for the AI Security integration:

  1. Create a new API key in your OpenAI tenant with read-only permissions.
  2. Send an email to support@openai.com with the following text:

Please add the “read” permissions of the Compliance API to the following key:
i. Last 4 characters of the API key: <LAST 4 DIGITS>
ii. Key name: <KEY NAME>
iii. Created by: <CREATED BY>

  1. Once OpenAI confirms the permissions for the API key, continue to configure the ChatGPT connector in the CMA.

Connecting ChatGPT Compliance API

To connect AI Security to ChatGPT Enterprise:

  1. From the CMA navigation menu, click AI Security > Integrations.

  2. In ChatGPT Compliance API, click Connect.
    ai_sec_chat_Gpt.png

  3. In the Add ChatGPT Compliance API panel, add the following details:

    • API Key
    • Workspace ID
    • Workspace Name

  4. Click Test Connection to verify the integration.

  5. Click Connect.