Repository Update: April 2025
A materially expanded, structured AI risk database gives APS governance and policy practitioners a single reference for comparing risk frameworks and identifying coverage gaps.
Key points
- MIT AI Risk Repository Version 3 now covers over 1,600 coded AI risks drawn from 65 published frameworks.
- Nine newly added frameworks include the final International Scientific Report on the Safety of Advanced AI and multi-agent risk taxonomies.
- APS risk and governance teams can use this as a structured reference to benchmark agency AI risk frameworks against global practice.
Implications for Australian agencies
- Consider APS agencies developing or reviewing AI risk frameworks could use the MIT repository to benchmark their own risk categorisation against the 65 published frameworks now indexed.
- Monitor Policy teams working on AI governance may want to monitor future repository updates, particularly as multi-agent and general-purpose AI risk taxonomies mature and potentially influence international standards.
Implications are AI-generated. Starting points, not advice — see methodology for how they're framed.
View original source
Copied.
"Repository Update: April 2025"
Source: MIT AI Risk Repository – Blog
Published: 1 April 2025
URL: https://airisk.mit.edu/blog/april-2025-update-of-the-ai-risk-repository-2
MIT's AI Risk Repository has released Version 3, expanding its structured database to over 1,600 coded AI risks from 65 published frameworks. The April 2025 update adds nine new frameworks including the final International Scientific Report on the Safety of Advanced AI, taxonomies of generative AI misuse, multi-agent risks, and systemic risks as defined under the EU AI Act. A new subdomain on multi-agent risks reflects growing attention to the governance challenges posed by interacting AI systems. The repository is freely accessible and designed to support researchers, policymakers, and practitioners in identifying and comparing risk frameworks.
Implications for Australian agencies:
- [Consider] APS agencies developing or reviewing AI risk frameworks could use the MIT repository to benchmark their own risk categorisation against the 65 published frameworks now indexed.
- [Monitor] Policy teams working on AI governance may want to monitor future repository updates, particularly as multi-agent and general-purpose AI risk taxonomies mature and potentially influence international standards.
Retrieved from SIMS, 18 July 2026.