Repository Updates: December 2024
A consolidated, taxonomy-structured AI risk database gives APS risk and governance practitioners a citable, evolving reference for risk categorisation work.
Key points
- MIT AI Risk Repository expanded to 13 new frameworks in December 2024, now covering over 1000 AI risks.
- Added frameworks include Australian-authored contributions and a UK Government Office for Science frontier AI risks report.
- Repository is structured living knowledge infrastructure - quarterly updates planned through 2025 for researchers and policymakers.
Summary
The MIT AI Risk Repository has released its December 2024 update, adding 13 new frameworks to a database now exceeding 1000 categorised AI risks drawn from 56 source documents. New additions include the International Scientific Report on the Safety of Advanced AI (Bengio et al.), the UK Government Office for Science frontier AI risks report, and academic work on generative AI misuse and long-term societal impacts. The repository uses dual taxonomies - causal and domain-based - and commits to quarterly updates through 2025. Authors represented include contributors from Australia, making this a resource with some direct relevance to Australian AI governance framing.
Implications for Australian agencies
- Monitor APS risk and governance teams may want to monitor the repository's 2025 quarterly updates as a source of evolving AI risk categorisation language useful for internal frameworks.
- Consider Agencies developing or reviewing AI risk registers could consider referencing the repository's domain taxonomy to align risk language with internationally recognised classifications.
Implications are AI-generated. Starting points, not advice.
"Repository Updates: December 2024" Source: MIT AI Risk Repository – Blog Published: 31 December 2024 URL: https://airisk.mit.edu/blog/framework-updates-december-2024 The MIT AI Risk Repository has released its December 2024 update, adding 13 new frameworks to a database now exceeding 1000 categorised AI risks drawn from 56 source documents. New additions include the International Scientific Report on the Safety of Advanced AI (Bengio et al.), the UK Government Office for Science frontier AI risks report, and academic work on generative AI misuse and long-term societal impacts. The repository uses dual taxonomies - causal and domain-based - and commits to quarterly updates through 2025. Authors represented include contributors from Australia, making this a resource with some direct relevance to Australian AI governance framing. Implications for Australian agencies: - [Monitor] APS risk and governance teams may want to monitor the repository's 2025 quarterly updates as a source of evolving AI risk categorisation language useful for internal frameworks. - [Consider] Agencies developing or reviewing AI risk registers could consider referencing the repository's domain taxonomy to align risk language with internationally recognised classifications. Retrieved from SIMS, 18 May 2026.