Download Framework
https://csrc.nist.gov/
nist-ai-100-2-adversarial-ml
NIST / AI 100-2 Adversarial ML icon

NIST / AI 100-2 Adversarial ML

Detailed taxonomy of adversarial attacks on ML systems covering attack types, attacker goals, and mitigations. The 2025 edition adds GenAI threats including prompt injection.
Type:

Standard

Domain:

Cybersecurity

Coverage:

Cybersecurity

Safety & Reputational Harm

Performance & Reliability

Region:

US

Tags:

GenAI

Content:

0 Risks

0 Controls

Version: 2024

Framework Definition

Risks and controls associated with the framework

Assessment Layer

Concrete evaluations linked to controls to assess pass or fail
No evaluation mapping defined yet.