Explore the Frameworks Behind the AI Risk Repository

4 Apr 2025 · MIT AI Risk Repository – Blog Global

A transparent, citable compilation of 65 AI risk frameworks gives APS governance teams a ready reference for grounding agency-level risk taxonomies.

Key points

Summary

MIT's AI Risk Repository has released a companion resource for Version 3 (April 2025) that publicly documents all 65 frameworks and documents underpinning its AI risk taxonomy. Presented as a navigable Google Slides deck, it includes key excerpts, diagrams, taxonomies, and citation details from academic, industry, and policy sources. The resource is designed to support researchers, policymakers, and governance professionals in examining source frameworks, building on existing taxonomies, and conducting their own audits. The underlying methodology is detailed in a peer-reviewed preprint on arXiv.

Implications for Australian agencies

Implications are AI-generated. Starting points, not advice.