The Security Implications of AI-Enabled Technology in the Military

Several world leaders and defence secretaries across the world have referred to a current ‘AI arms race’ in which militaries vie for technological superiority across artificial intelligence-enabled technologies. Concurrently, active research highlights a significant range of security risks, misuse cases, and concerns on the ethics and strategic implications of using AI-enabled technologies without sufficient validation and oversight. My research focuses on exploring this tension within the military AI innovation landscape, in particular within the UK and the US. Through highlighting the technical, institutional, market and strategic dynamics of the design and deployment of AI-enabled technology for defence I map out likely unforeseen or unintended security consequences of current practices. My research also looks at potential options to mitigate the dynamics most likely to lead to 'irresponsible' AI including the role of international norms as well as industry frameworks. The 3M presentation will conclude by drawing the attention of listener's to a 'so what'? Military innovation has historically fed into the civic space (with drones or facial recognition being more recent examples) and so the norms and constraints on emerging technologies at this stage set a precedent for future deployment. By offering an overview of current gaps and mitigations in the field, I hope to communicate the importance of effective norms-setting as an urgent task to prevent the proliferation of technology with a net negative impact.