When AI Fails, What Works? A Data-Driven Taxonomy of Real-World AI Risk Mitigation Strategies

TL;DR

This paper develops an empirically grounded taxonomy of real-world AI risk mitigation strategies by analyzing thousands of incident reports, identifying new response categories, and enhancing system monitoring to prevent failures.

Contribution

It introduces four new mitigation categories to the existing taxonomy, based on analysis of nearly 9,700 incident reports, expanding coverage of AI failure responses.

Findings

Identified 9,629 mitigation actions, including 67% new patterns.

Expanded taxonomy with four new mitigation categories.

Enhanced post-deployment monitoring for systemic failure prevention.

Abstract

Large language models (LLMs) are increasingly embedded in high-stakes workflows, where failures propagate beyond isolated model errors into systemic breakdowns that can lead to legal exposure, reputational damage, and material financial losses. Building on this shift from model-centric risks to end-to-end system vulnerabilities, we analyze real-world AI incident reporting and mitigation actions to derive an empirically grounded taxonomy that links failure dynamics to actionable interventions. Using a unified corpus of 9,705 media-reported AI incident articles, we extract explicit mitigation actions from 6,893 texts via structured prompting and then systematically classify responses to extend MIT's AI Risk Mitigation Taxonomy. Our taxonomy introduces four new mitigation categories, including 1) Corrective and Restrictive Actions, 2) Legal/Regulatory and Enforcement Actions, 3) Financial, Economic, and Market Controls, and 4) Avoidance and Denial, capturing response patterns that are becoming increasingly prevalent as AI deployment and regulation evolve. Quantitatively, we label the mitigation dataset with 32 distinct labels, producing 23,994 label assignments; 9,629 of these reflect previously unseen mitigation patterns, yielding a 67% increase of the original subcategory coverage and substantially enhancing the taxonomy's applicability to emerging systemic failure modes. By structuring incident responses, the paper strengthens "diagnosis-to-prescription" guidance and advances continuous, taxonomy-aligned post-deployment monitoring to prevent cascading incidents and downstream impact.