A pragmatic classification framework for AI incident monitoring

TL;DR

This paper introduces a structured framework for monitoring AI incidents over time, accounting for reporting biases and system exposure, to improve governance and safety in high-reliability sectors.

Contribution

It presents a novel methodological framework combining structured questions, tiered estimation, and classification to analyze AI incident trends with limited data.

Findings

Framework clarifies incident trend analysis despite data limitations.

Case studies demonstrate governance insights from the framework.

Proof of concept shows practical utility for AI safety monitoring.

Abstract

Incident monitoring can drive safety improvements in high-reliability industries and population-scale technologies, but remains underdeveloped in AI governance. Public databases catalog thousands of AI incidents, but simple incident counts conflate media reporting propensity, system deployment ("exposure"), and harm frequency per unit exposure. We propose a methodological framework that accounts for these factors and calibrates confidence to available evidence in analyzing how AI incidents change over time. The framework comprises three components: a structured monitoring question that defines the scope of the analysis; a tiered estimation process that separately derives harm and exposure trends, including through LLM-assisted filtering of public incident databases; and a classification scheme that maps the resulting trend estimates onto actionable governance categories (Escalating, Mitigating, Concentrating, Receding or Unclassifiable). Through case studies, we examine the framework's clarifying power and limitations, demonstrate governance insight despite real-world data constraints, and provide a proof of concept for AI incident monitoring as a practical governance tool.