Bigger, Faster, Better? Techno-Optimism and Conflict Anticipation

• In Event: Emerging Technologies and International Security

Abstract

In light of the rapid advancements in artificial intelligence (AI) in recent years, numerous organizations, governments, think tanks, and private companies are rushing to develop and implement AI-powered tools to inform conflict forecasting. AI, so the hope, can cope more effectively and accurately with today’s complex conflict landscape. These efforts are often infused with technological optimism, premised on the belief that what actors lack is better knowledge, rather than resources, power, or political will. At the same time, the subfield of quantitative conflict studies within Political Science has long made use of AI to predict outbreaks of armed conflict by employing machine learning (ML) algorithms. However, this type of research is primarily focused on the technicalities of improving the modeling and does not usually discuss policy implications and applications, nor does it critically assess the often-unspoken assumptions underwriting many efforts at integrating AI into conflict anticipation: that what we lack is better knowledge and faster assessments—and that intelligent technologies can get us there.

Leveraging original data from expert interviews with policymakers, practitioners, and conflict modelers across various sectors, this paper examines how AI shapes the production of knowledge geared at anticipating and developing policies to prevent armed conflict. Through the close and critical examination of the underlying assumptions, expectations, and aspirations that drive the push for integrating AI into conflict analytics, this research contributes to current scholarship by demystifying the current ‘AI hype’ and drawing out how it works to transform questions of power into questions of process, and problems of politics into problems of technological nature.

Author

• Johanna Rodehau-Noack, Stanford University