Impact (9 techniques)
MITRE ATLAS tactic
The adversary is trying to manipulate, interrupt, erode confidence in, or destroy your AI systems and data. Impact consists of techniques that adversaries use to disrupt availability or compromise integrity by manipulating business and operational processes. Techniques used for impact can include destroying or tampering with data. In some cases, business processes can look fine, but may have been altered to benefit the adversaries' goals. These techniques might be used by adversaries to follow through on their end goal or to provide cover for a confidentiality breach.
Techniques
- AML.T0015 - Evade AI ModelMaturity: realized
- AML.T0029 - Denial of AI ServiceMaturity: demonstrated
- AML.T0031 - Erode AI Model IntegrityMaturity: realized
- AML.T0034 - Cost HarvestingMaturity: feasible
- AML.T0046 - Spamming AI System with Chaff DataMaturity: feasible
- AML.T0048 - External HarmsMaturity: realized
- AML.T0059 - Erode Dataset IntegrityMaturity: demonstrated
- AML.T0101 - Data Destruction via AI Agent Tool InvocationMaturity: realized
- AML.T0112 - Machine CompromiseMaturity: demonstrated
AI red teaming training, taught by practitioners.
Hands-on courses on adversarial AI, prompt injection, and AI security operations.
View AI Security Courses