Safety and War: Safety and Security Assurance of Military AI Systems
Safety assurance frameworks for military AI are an emerging gap in global governance - Australian Defence and DISR policy teams may encounter these arguments in multilateral forums.
Key points
- AI Now Institute argues military AI systems like Lavender and Gospel lack safety assurance, oversight, and accountability.
- The paper calls for safety engineering frameworks applied to military AI - directly relevant to defence AI governance debates.
- This is introductory framing for a future research series; substantive technical guidance is not yet published.
Implications for Australian agencies
- Monitor Defence, DISR, and DFAT policy teams may want to monitor this research series as it develops more specific safety assurance methodology applicable to multilateral AI governance discussions.
- Consider Agencies working on AI risk frameworks could consider how safety engineering approaches advocated here compare with existing Australian Government AI assurance guidance for high-stakes contexts.
Implications are AI-generated. Starting points, not advice — see methodology for how they're framed.
View original source
Copied.
"Safety and War: Safety and Security Assurance of Military AI Systems"
Source: AI Now Institute – Publications
Published: 25 June 2024
URL: https://ainowinstitute.org/publications/safety-and-war-safety-and-security-assurance-of-military-ai-systems
The AI Now Institute outlines a research agenda examining whether military AI systems can meet safety and security assurance standards required for safety-critical applications. The paper critiques systems like Lavender and Gospel deployed in Gaza, arguing their lack of independent evaluation, opaque development, and absence of enforceable risk thresholds represent a fundamental governance failure. It draws on safety engineering concepts to argue that current AI armaments cannot be validated without defined use cases and risk thresholds, and tentatively supports a moratorium on military AI deployment until rigorous assurance frameworks exist. Future instalments will address safety methodology, export controls, and supply chain vulnerabilities.
Implications for Australian agencies:
- [Monitor] Defence, DISR, and DFAT policy teams may want to monitor this research series as it develops more specific safety assurance methodology applicable to multilateral AI governance discussions.
- [Consider] Agencies working on AI risk frameworks could consider how safety engineering approaches advocated here compare with existing Australian Government AI assurance guidance for high-stakes contexts.
Retrieved from SIMS, 18 July 2026.