Australia’s Department of Defence recently released a new AI policy establishing three requirements for military AI: legal compliance, individual accountability with explainability, and proportionate risk controls including testing and evaluation. The policy covers all AI types and uniquely mandates Article 36 legal reviews for weapon systems, but it lacks implementation roadmaps, resourcing, and detailed testing procedures. The document aligns with US and UK principles yet emphasizes national frameworks’ growing importance.
Originally Appeared Here
