MIT AI Risk Repository Selected for 2025 Paris Peace Forum AI Action Summit
A consolidated, publicly accessible AI risk database drawn from 56 frameworks offers APS risk and governance teams a ready-made comparative reference.
Key points
- MIT AI Risk Repository was selected among 50 projects from 770 applications to present at the 2025 Paris AI Action Summit.
- The repository consolidates over 1,000 AI risks from 56 frameworks into a publicly accessible, categorised database with two taxonomies.
- The resource has over 100,000 visits and is already adopted by companies, governments, and researchers - a ready-made reference for APS risk work.
Summary
The MIT AI Risk Repository has been selected to present at the 2025 Paris Peace Forum AI Action Summit, recognising its role in systematising AI risk categorisation. The repository draws on 56 existing AI risk frameworks to compile over 1,000 discrete risks, structured through a Causal Taxonomy and a Domain Taxonomy spanning 7 domains and 23 subdomains. With broad international adoption by governments, companies, and researchers, it represents a useful baseline resource for agencies conducting AI risk assessments or developing their own risk frameworks.
Implications for Australian agencies
- Consider Agencies developing or reviewing AI risk frameworks could assess whether the MIT AI Risk Repository's taxonomy aligns with or usefully supplements existing Australian frameworks such as the Responsible AI framework or NIST AI RMF mappings.
- Monitor APS practitioners may want to monitor outputs from the AI Action Summit for emerging international consensus on AI risk categorisation that could inform future Australian standards or guidance.
Implications are AI-generated. Starting points, not advice.
"MIT AI Risk Repository Selected for 2025 Paris Peace Forum AI Action Summit" Source: MIT AI Risk Repository – Blog Published: 17 January 2025 URL: https://airisk.mit.edu/blog/mit-ai-risk-repository-selected-for-2025-paris-peace-forum-ai-action-summit The MIT AI Risk Repository has been selected to present at the 2025 Paris Peace Forum AI Action Summit, recognising its role in systematising AI risk categorisation. The repository draws on 56 existing AI risk frameworks to compile over 1,000 discrete risks, structured through a Causal Taxonomy and a Domain Taxonomy spanning 7 domains and 23 subdomains. With broad international adoption by governments, companies, and researchers, it represents a useful baseline resource for agencies conducting AI risk assessments or developing their own risk frameworks. Implications for Australian agencies: - [Consider] Agencies developing or reviewing AI risk frameworks could assess whether the MIT AI Risk Repository's taxonomy aligns with or usefully supplements existing Australian frameworks such as the Responsible AI framework or NIST AI RMF mappings. - [Monitor] APS practitioners may want to monitor outputs from the AI Action Summit for emerging international consensus on AI risk categorisation that could inform future Australian standards or guidance. Retrieved from SIMS, 18 May 2026.