Record summary
A quick snapshot of what this page covers.
Control summary
What this defense is meant to help prevent.
- ATLAS ID
- AML.M0006
- Priority score
- 55
Covered techniques
Attacks this defense is designed to help with.
AML.T0010.001 - AI Software
Using multiple different models ensures minimal performance loss if security flaw is found in tool for one model or family.
AML.T0043.001 - Black-Box Optimization
Using an ensemble of models increases the difficulty of crafting effective adversarial data and improves overall robustness.
AML.T0043.002 - Black-Box Transfer
Using an ensemble of models increases the difficulty of crafting effective adversarial data and improves overall robustness.
AML.T0043 - Craft Adversarial Data
Using an ensemble of models increases the difficulty of crafting effective adversarial data and improves overall robustness.
AML.T0014 - Discover AI Model Family
Use multiple different models to fool adversaries of which type of model is used and how the model used.
AML.T0031 - Erode AI Model Integrity
Using multiple different models increases robustness to attack.
AML.T0015 - Evade AI Model
Using multiple different models increases robustness to attack.
AML.T0043.004 - Insert Backdoor Trigger
Using an ensemble of models increases the difficulty of crafting effective adversarial data and improves overall robustness.
AML.T0043.003 - Manual Modification
Using an ensemble of models increases the difficulty of crafting effective adversarial data and improves overall robustness.
AML.T0010.003 - Model
Using multiple different models ensures minimal performance loss if security flaw is found in tool for one model or family.
AML.T0043.000 - White-Box Optimization
Using an ensemble of models increases the difficulty of crafting effective adversarial data and improves overall robustness.
Source
Where this page information comes from.
Original source
Original source links
Open the public records and source datasets used for this page.