Incident Tracker - June 2025 Update

18 Jul 2025 · MIT AI Risk Repository – Blog Global

A structured, updated AI incident database with national security and harm severity frameworks offers APS risk and governance teams a ready reference for evidence-based risk assessment.

Key points

Summary

The MIT AI Risk Repository has released a June 2025 update to its AI incident tracker, now synchronised with the AI Incident Database through incident #1116. Key enhancements include a simplified 1-5 harm severity scale, a national security impact assessment covering physical security, information warfare, sovereignty, economic security, and societal stability, and Ishikawa diagrams to surface potential causal factors. Spider-chart visualisations allow comparative profiling of incidents. The update also adds AI system goal classification aligned to a recognised taxonomy.

Implications for Australian agencies

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