MITRE ATLASsub-technique
AML.T0043.001 Black-Box Optimization
- Tactics
- Ai Attack Staging
- Created
- 2021-05-13
Description
In Black-Box attacks, the adversary has black-box (i.e. AI Model Inference API Access via API access) access to the target model.
With black-box attacks, the adversary may be using an API that the victim is monitoring.
These attacks are generally less effective and require more inferences than White-Box Optimization attacks, but they require much less access.
Related
Subtechnique of · MITRE ATLAS (1)
- AML.T0043 Craft Adversarial Data technique
Tactic · MITRE ATLAS (1)
- AML.TA0001 AI Attack Staging tactic
Referenced by MITRE ATLAS · mitigates (7)
- AML.M0019 Control Access to AI Models and Data in Production mitigation
- AML.M0002 Passive AI Output Obfuscation mitigation
- AML.M0003 Model Hardening mitigation
- AML.M0004 Restrict Number of AI Model Queries mitigation
- AML.M0006 Use Ensemble Methods mitigation
- AML.M0010 Input Restoration mitigation
- AML.M0015 Adversarial Input Detection mitigation