MITRE ATLAStechnique

AML.T0018 Manipulate AI Model

Published 2021-05-13 Modified 2025-04-14 Official page ↗
Tactics
Persistence, Ai Attack Staging
Created
2021-05-13

Description

Adversaries may directly manipulate an AI model to change its behavior or introduce malicious code. Manipulating a model gives the adversary a persistent change in the system. This can include poisoning the model by changing its weights, modifying the model architecture to change its behavior, and embedding malware which may be executed when the model is loaded.

Related

Tactic · MITRE ATLAS (2)

Referenced by MITRE ATLAS · mitigates (3)

  • AML.M0005 Control Access to AI Models and Data at Rest mitigation
  • AML.M0008 Validate AI Model mitigation
  • AML.M0013 Code Signing mitigation

Referenced by MITRE ATLAS · subtechnique of (3)