MITRE ATLAS
EvalsDatabase / taxonomy (not an eval — tooling/reference entry)
ATT&CK-style matrix of adversary tactics against AI systems, plus case studies.
- Category
- Risk taxonomies, incident data & standards
- Insurance exposure
- Breach/cyber
- Publisher
- MITRE
- Publisher type
- Government / standards body
- License
- Apache-2.0
- Status
- Active (<90d)
- Last commit
- 2026-09-15
- Stars
- 199
- Notes
- Standard vocabulary for AI attack TTPs.
Wanich Risk Lab, 2026.