New AI Now Paper Highlights Risks of Commercial AI Used In Military Contexts
Signals a growing policy critique of commercial AI in defence contexts - relevant to Australian agencies navigating dual-use AI procurement and national security risk frameworks.
Key points
- Foundation models integrated into military intelligence, surveillance, and targeting systems pose significant and underappreciated national security risks.
- The paper argues CBRN proliferation concerns dominate policy debate while deployed AI targeting systems in Gaza present more immediate dangers.
- Recommendations to insulate military AI from commercial foundation models have indirect relevance to Australian Defence and intelligence AI procurement policy.
Summary
This AI Now Institute paper examines the national security risks arising from the use of commercial AI foundation models in military intelligence, surveillance, and targeting contexts. It argues that existing policy responses - such as compute thresholds and export controls - are insufficient, and that the reliance on personal data within commercial AI systems creates exploitable vulnerabilities. The paper recommends segregating military AI from commercial foundation models and treating personal data use in such systems as a distinct national security concern. It draws on examples including Israeli targeting systems deployed in Gaza.
Implications for Australian agencies
- Monitor Defence, DISR, and national security-adjacent policy teams may want to monitor this emerging critique as it could influence allied-nation AI procurement and governance norms.
- Consider Agencies involved in AI procurement or dual-use AI governance could consider whether the paper's framing of commercial-AI-in-military-contexts risks surfaces gaps in current Australian policy guidance.
Implications are AI-generated. Starting points, not advice.
"New AI Now Paper Highlights Risks of Commercial AI Used In Military Contexts" Source: AI Now Institute – Publications Published: 22 October 2024 URL: https://ainowinstitute.org/publications/new-ai-now-paper-highlights-risks-of-commercial-ai-used-in-military-contexts This AI Now Institute paper examines the national security risks arising from the use of commercial AI foundation models in military intelligence, surveillance, and targeting contexts. It argues that existing policy responses - such as compute thresholds and export controls - are insufficient, and that the reliance on personal data within commercial AI systems creates exploitable vulnerabilities. The paper recommends segregating military AI from commercial foundation models and treating personal data use in such systems as a distinct national security concern. It draws on examples including Israeli targeting systems deployed in Gaza. Implications for Australian agencies: - [Monitor] Defence, DISR, and national security-adjacent policy teams may want to monitor this emerging critique as it could influence allied-nation AI procurement and governance norms. - [Consider] Agencies involved in AI procurement or dual-use AI governance could consider whether the paper's framing of commercial-AI-in-military-contexts risks surfaces gaps in current Australian policy guidance. Retrieved from SIMS, 18 May 2026.