Analytics Low

AI model discovery

A cloud identity listed available AI models. This behavior often suggests reconnaissance on AI models and potential misuse. MITRE ATLAS Technique: AML.T0007 - Discover ML Artifacts.

Module:
Cortex Cloud
Licensed by:
Cloud Runtime Security (CRS)
Data source:
AWS Audit Log, Azure Audit Log, Gcp Audit Log
ATT&CK tactics: Discovery (TA0007)
ATT&CK techniques: Cloud Service Discovery (T1526)
Detector tags: Cloud AI Infrastructure Analytics
Attacker's goals:

Gather information about AI models in the environment.

Investigative actions:

Determine which AI models were enumerated. Monitor the usage of affected AI models to detect potential misuse, such as unusual or excessive access attempts. Investigate any unusual activity originating from the suspected identity.

Test period:
10 Minutes
Deduplication:
5 Days