With Antonia Toffert
A new systematic review by researchers from the OII’s Digital Ethics and Defence Technologies Research Group, comprising Antonia Toffert, Katharina Klotz, Huw Roberts, Cyril Birks and Professor Mariarosaria Taddeo, explains why.
Drawing on an analysis of over 1,000 publications, the researchers find that the answer lies less in any single failing than in how the obstacles interact.
They identify seven barriers to the ethical governance of AI in defence, spanning governance and structural challenges, conceptual ambiguity, strategic constraints, and shortages of skills and dedicated funding.
They explain how the deeper problem is not the barriers in isolation, but the way they reinforce one another into systemic governance failure. Fix one, and the pressure simply shifts to the next. The paper sets out three priorities for closing the implementation gap:
1️⃣ Build more coherent and interoperable governance architectures
2️⃣ Invest in institutional capacity through dedicated resources and interdisciplinary expertise
3️⃣ Embed ethical oversight across the whole AI lifecycle, rather than bolting it on as a compliance retrofit
👉 Building effective AI governance in defence requires more than tackling barriers individually. It requires a systemic approach.
Read the full paper ‘Seven barriers to the ethical governance of artificial intelligence in defence’, published in the journal AI and Society: https://rdcu.be/fmM3X
