Tools & databases // the current ecosystem

• AI Incident Database (AIID): open-source repository of documented AI incidents and harms.
• OECD AI Incidents Monitor (AIM): curated collection of AI-related incidents with standardized metadata and cross-jurisdictional coverage.
• AVID: framework and database for documenting AI vulnerabilities, failures, and associated risks.
• AIAAIC: database of AI incidents, controversies, and societal impacts compiled from public sources.
• MITRE ATLAS: knowledge base of adversarial tactics, techniques, and mitigations targeting AI systems.
• Agent observability platforms (LangSmith, Langfuse, Arize Phoenix, W&B Weave): tools for tracing, evaluating, and debugging LLM and agent workflows.