Revisiting The ‘Stability–Instability Paradox’ In AI-Enabled Warfare: A Modern-Day Promethean Tragedy Under The Nuclear Shadow?

In an article for Review of International Studies, James Johnson examines the stability-instability paradox: the idea that nuclear weapons may deter large-scale war while increasing the likelihood of lower-level conflict between nuclear-armed rivals. It challenges–and partially inverts–the dominant interpretation of the paradox, which assumes that mutual vulnerability creates a reliable ceiling on escalation and reduces the risks associated with military actions below the nuclear threshold.

The article argues that insufficient attention has been paid to the possibility that conflicts below the nuclear threshold can escalate in unintended ways. Using empirically grounded fictional scenarios, it explores how emerging forms of AI-enabled warfare may interact with escalation dynamics between nuclear-armed states. Rather than predicting the future, the scenarios test and interrogate key assumptions embedded in prevailing theories of strategic stability.

The central argument is that the apparent strategic cap provided by nuclear deterrence may be more fragile than commonly assumed. As AI systems become increasingly integrated into military operations, low-level aggression between nuclear powers could generate new pathways to misperception, inadvertent escalation, and unintended nuclear use.