ai-incidents

Single source of truth for GenAI and agentic AI security incidents, mapped to OWASP LLM Top 10, OWASP Agentic Top 10 (ASI), NIST AI RMF, and MITRE ATLAS.

An open database of real-world AI agent incidents.

5,660 AI security links: prompt injection, LLM & AI agent security, MCP, coding agents, agent skills & supply chain, red teaming, guardrails, sandboxing, incidents & CVEs, governance. 570 arXiv papers, 1,155 GitHub repos, 570 X posts, 29 categories of plain markdown you can grep, diff and fork. Obsidian vault export. Added, not endorsed.

Public registry for AI code failures. AICI identifiers. Detection rule mapping. Vendor notification.

A sourced library of real-world AI failures and what to learn from them.

PipeRoll - Agent Incident Registry: verified public records of AI-agent failures. Open data, open submissions, verified registration.

Replication package for a preliminary 100-incident AI Incident Database audit of public-record reconstructability.

MVP pipeline for LLM-assisted annotation of AI incidents from the AIID. Imports AIID backup data, runs GPT structured-output classification against GMF taxonomy categories, and compares predictions to human gold labels via a Streamlit UI. FastAPI + SQLAlchemy + Pydantic v2.

Replication materials for "Category, Not Severity: How Large Language Models Misclassify AI Incidents Under the EU AI Act Risk Taxonomy" (SiKDD 2026) - annotation rubric, prompts, expert labels, the classification harness, and 4,198 model classifications.

Public AI incident atlas for governance, source review, evidence readiness, and public deployment-safe incident records.

Curated incidents, standards, and regulatory context on financial governance for autonomous AI agents. Updated as the space evolves.

Real AI agent incident post-mortems: Gateway outages, model failures, runaway tasks, and cascading errors. 33 operational lessons from production multi-agent systems.

Landing site for the AI Incident Card spec. Live at incidents.kineticgain.com.

Walks the Kinetic Gain Protocol Suite graph from an AI Incident Card and emits a structured remediation plan. Rust + petgraph; small surface, composes with the rest of the portfolio.

In-depth reconstructions of incidents where an AI system's autonomy was central to real-world security harm, built from primary sources.

Curated incidents, standards, and thinking on financial governance for autonomous AI agents

Sourced, bilingual (NL/EN) dataset of AI incidents, harms and politics in 2026, with severity and disclosure delay. CC BY 4.0.