Defense Evasion (15 techniques)
MITRE ATLAS tactic
The adversary is trying to avoid being detected by AI-enabled security software. Defense Evasion consists of techniques that adversaries use to avoid detection throughout their compromise. Techniques used for defense evasion include evading AI-enabled security software such as malware detectors.
Techniques
- AML.T0015 — Evade AI ModelMaturity: realized
- AML.T0054 — LLM JailbreakMaturity: demonstrated
- AML.T0067 — LLM Trusted Output Components ManipulationMaturity: demonstrated
- AML.T0068 — LLM Prompt ObfuscationMaturity: demonstrated
- AML.T0071 — False RAG Entry InjectionMaturity: demonstrated
- AML.T0073 — ImpersonationMaturity: realized
- AML.T0074 — MasqueradingMaturity: realized
- AML.T0076 — Corrupt AI ModelMaturity: realized
- AML.T0081 — Modify AI Agent ConfigurationMaturity: demonstrated
- AML.T0092 — Manipulate User LLM Chat HistoryMaturity: demonstrated
- AML.T0094 — Delay Execution of LLM InstructionsMaturity: demonstrated
- AML.T0097 — Virtualization/Sandbox EvasionMaturity: realized
- AML.T0107 — Exploitation for Defense EvasionMaturity: demonstrated
- AML.T0109 — AI Supply Chain Rug PullMaturity: realized
- AML.T0111 — AI Supply Chain Reputation InflationMaturity: demonstrated
AI red teaming training, taught by practitioners.
Hands-on courses on adversarial AI, prompt injection, and AI security operations.
View AI Security Courses