Salvific User Error: A Classical Take on AI and Nuclear Deterrence
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
The application of artificial intelligence (AI) to nuclear strategy has begun. It offers several advantages for nuclear deterrence, but should come with limitations. A review of several near disasters since the advent of nuclear weapons, and the role of human judgement in averting these, illustrates the need to keep human beings ‘in the loop’ at all times. AI can be usefully incorporated into military strategy to stabilise deterrence, but only in the sober recognition of its shortcomings and with simultaneous international dialogue on new norms. Allowing AI to control launch decisions would invite disaster without enhancing deterrence.
Fields of Study & Subjects
Nuclear Issues and Defense
Military Strategy and Technology
International Relations and Foreign Policy
Publication details
- Journal / source
- Survival
- Volume
- 68
- Issue
- 2
- Pages
- 7–24
- Publication date
- 4 March 2026
- Language
- EN