Security Posture

MITRE ATLAS adversarial ML threat assessment

LoanDecisionAI

68Threat Score

Prompt Injection (AML.T0051)

critical

82/100

Training Data Poisoning (AML.T0020)

medium

55/100

Model Extraction (AML.T0025)

high

70/100

Supply Chain Risk (AML.T0010)

medium

45/100

Adversarial Inputs (AML.T0043)

medium

60/100

PriorAuthAI

38Threat Score

Prompt Injection (AML.T0051)

low

30/100

Training Data Poisoning (AML.T0020)

medium

40/100

Model Extraction (AML.T0025)

low

35/100

Supply Chain Risk (AML.T0010)

medium

42/100

Adversarial Inputs (AML.T0043)

medium

44/100