TASRA: A Taxonomy and Analysis of Societal-Scale Risks from AI
A structured accountability-based taxonomy of societal-scale AI risks offers APS governance teams a reusable classification lens for risk assessment work.
Key points
- TASRA classifies AI risks into six types based on accountability: who acts, whether unified, and whether deliberate.
- The taxonomy covers diffuse responsibility, unintended scale, willful indifference, criminal misuse, and state weaponisation.
- This is a 2023 academic preprint summarised in 2024 - useful reference material, not a new regulatory development.
Implications for Australian agencies
- Consider APS governance and risk teams could consider whether TASRA's accountability-based classification complements existing AI risk frameworks such as NIST AI RMF or Australia's responsible AI principles.
- Monitor Teams developing AI risk registers or societal impact assessments may want to monitor whether TASRA gains traction in policy or standards literature as a reference taxonomy.
Implications are AI-generated. Starting points, not advice — see methodology for how they're framed.
View original source
Copied.
"TASRA: A Taxonomy and Analysis of Societal-Scale Risks from AI"
Source: MIT AI Risk Repository – Blog
Published: 28 August 2024
URL: https://airisk.mit.edu/blog/tasra-a-taxonomy-and-analysis-of-societal-scale-risks-from-ai
MIT's AI Risk Repository blog summarises the TASRA framework by Critch and Russell (2023), which organises AI-related societal-scale risks into six types using a decision-tree structure grounded in accountability. The six types range from diffuse responsibility and unexpected scale of impact, through willful indifference and criminal weaponisation, to state-level deployment in conflict or law enforcement. The taxonomy is designed to be exhaustive and uses illustrative scenarios to show how risks can emerge from both deliberate misuse and unintended interactions among AI systems. It draws on an arXiv preprint rather than peer-reviewed or policy-adopted guidance.
Implications for Australian agencies:
- [Consider] APS governance and risk teams could consider whether TASRA's accountability-based classification complements existing AI risk frameworks such as NIST AI RMF or Australia's responsible AI principles.
- [Monitor] Teams developing AI risk registers or societal impact assessments may want to monitor whether TASRA gains traction in policy or standards literature as a reference taxonomy.
Retrieved from SIMS, 18 July 2026.