Resources
Reference material to read alongside this watch: the frameworks that name the techniques, and the databases that record the incidents.
Frameworks and taxonomies
- MITRE ATLAS
Adversary tactics and techniques against AI systems, with case studies. The ATT&CK counterpart for machine learning.
- OWASP GenAI Security Project
Top 10 for LLM applications, the agentic risk list, and a quarterly exploit round-up.
- NIST AI Risk Management Framework
Risk framework and companion documents, including the generative AI profile.
Incident databases
- AI Incident Database
Community-indexed record of harms and near misses involving deployed AI systems.
- AI Vulnerability Database (AVID)
Open taxonomy and database of failure modes and vulnerabilities in AI models.
- OECD AI Incidents and Hazards Monitor
Global monitor of AI incidents built from news reporting.
- MIT AI Risk Repository
Living database of AI risks extracted from published frameworks.