Azure AI API keys listed from previously unseen application

This rule is part of a beta feature. To learn more, contact Support.
azure

Classification:

attack

Tactic:

TA0006-credential-access

Technique:

T1555-credentials-from-password-stores

Set up the azure integration.

This page is not yet available in Spanish. We are working on its translation.
If you have any questions or feedback about our current translation project, feel free to reach out to us!

Goal

Detect when a previously unseen application lists keys for Azure AI and ML services.

Strategy

Monitor Azure activity logs for when a previously unseen application lists keys for Azure AI and ML services. This may indicate an attacker listing keys with compromised Azure tokens through the Azure CLI, direct web requests, or other tooling. Compromised keys may be used in attacks such as LLM hijacking (LLMjacking).

Triage and response

  1. Verify if the user or service principal is expected to be working with Azure AI and ML services, and if they are using these services outside of available web portals.
    • Use the value {{@identity.claims.appid}} in the Enterprise Application section of Microsoft Entra to gather further context about the application used.
  2. If the activity is not expected, investigate activity around the user and IP generating this event.