MITRE ATLAS
Also called Adversarial Threat Landscape for Artificial-Intelligence Systems.
MITRE's ATT&CK-style knowledge base of tactics and techniques used against AI systems.
What it is
A public catalog, maintained by MITRE, of adversary tactics, techniques and case studies targeting machine learning and AI. It uses the same tactic and technique structure as ATT&CK, with IDs that start with AML, such as AML.T0051 for LLM prompt injection.
Why it matters
It gives AI threats a shared vocabulary. Detection, red team findings and reports can name a technique precisely instead of describing "an AI attack".
In practice
A red team report maps each finding to an ATLAS technique ID, so the defenders can line the results up against their own coverage map.
What to look for
- Technique IDs in the form AML.Txxxx in reports and detections.
- Case studies that match the kind of AI system you run.
Common mistake
Using it as a checklist of everything that can go wrong. It records what has been observed and studied, so pair it with your own threat model.
Practice it nearby
CyberKit has no tool for this term itself. These are the closest places to practice it.
In the field
Well-known tools where this shows up on the job. They are listed to show the landscape, not as a recommendation.
-
garak Open source
Probes language models for prompt injection, jailbreaks and data leakage.
-
PyRIT Open source
Microsoft's framework for red teaming generative AI systems.
-
ModelScan Open source
Scans model files for code that would run when the model is loaded.
Read next
- MITRE ATT&CK: A public knowledge base of real adversary tactics and techniques.
- AI Red Teaming: Attacking an AI system on purpose, before someone else does, to find how it fails.
- Prompt Injection: Text crafted to make a language model ignore its instructions and follow the attacker's instead.
- Data Poisoning: Tampering with the data a model learns from so that it learns the wrong thing.