Delegating strategic decision-making to machines: Dr. Strangelove Redux?

Abstract

ABSTRACT Will the use of artificial intelligence (AI) in strategic decision-making be stabilizing or destabilizing? What are the risks and trade-offs of pre-delegating military force to machines? How might non-nuclear state and non-state actors leverage AI to put pressure on nuclear states? This article analyzes the impact of strategic stability of the use of AI in the strategic decision-making process, in particular, the risks and trade-offs of pre-delegating military force (or automating escalation) to machines. It argues that AI-enabled decision support tools - by substituting the role of human critical thinking, empathy, creativity, and intuition in the strategic decision-making process - will be fundamentally destabilizing if defense planners come to view AI’s ‘support’ function as a panacea for the cognitive fallibilities of human analysis and decision-making. The article also considers the nefarious use of AIenhanced fake news, deepfakes, bots, and other forms of social media by non-state actors and state proxy actors, which might cause states to exaggerate a threat from ambiguous or manipulated information, increasing instability.