Regulating autonomous and AI-enabled weapon systems: the dangers of hype

Abstract

In many debates surrounding autonomous weapon systems (AWS) or AI-enabled platforms in the military, critics present both over- and under-hyped presentations of the capabilities of such systems, creating a risk of derailing critical debates on how best to regulate these in the military. In particular, in this article, I show that critics utilize over-hype to generate fear about the capabilities of such systems or to create objections that do not hold for more realistically viewed platforms, and they use under-hype to sell AWS and military AI short, creating an image of these as far less capable than is in actuality the case. The hyped presentations in this debate also gloss over many core realities of how modern militaries function, what sorts of platforms they are seeking to develop and use, and what actual combatants are likely to be willing to deploy in real warfighting scenarios. More critically for the regulatory debates themselves, hype (both over and under) forces genuine but subtle arguments on issues with autonomous and AI-enabled systems to be sidelined as scholars deal with the more politically divisive topics brought to the fore by critics. Finally, over- and under-hype creates grave risks of skewing the regulatory debates far enough from the realities of AWS and military AI development and deployment that central state actors may lose willingness to support any eventual treaties established. Thus, in their fervor to generate objections and force rapid regulation of AWS and military AI, critics risk alienating those key players most necessary for such regulation to be globally meaningful and effective.