Artificial Intelligence and Nuclear Weapons Proliferation: The Technological Arms Race for (In)visibility

Area of Science:

• International Security
• Technology Policy
• Arms Control

Background:

• The global nonproliferation regime has limited nuclear weapon states to nine.
• Emerging technologies are significantly altering nuclear risks and proliferation dynamics.
• An intensifying technological arms race between proliferation-enabling technologies (PETs) and detection-enhancing technologies (DETs) is underway.

Purpose of the Study:

• To analyze the impact of technological innovation on nuclear proliferation.
• To model the strategic interplay between PETs and DETs, particularly with the rise of AI.
• To evaluate the effectiveness of detection strategies against advancing proliferation capabilities.

Main Methods:

• Development of a formal model using a Relative Advantage Index (RAI) to quantify the PETs vs. DETs balance.
• Scenario-based simulations to assess cumulative nuclear breakout risk under varying technological growth rates.
• Analysis of asymmetric technological advancement, including AI-driven PET growth and stepwise DET improvements.

Main Results:

• Asymmetric technological advancement, especially AI-accelerated PETs, expands uncertainty in proliferation detectability.
• The pace of innovation in PETs and DETs increasingly shapes proliferation patterns.
• Traditional monitoring and verification methods are challenged by AI's capabilities.

Conclusions:

• Detection alone may become insufficient without broader governance of PETs.
• A strategic shift is needed, anticipating future technological advancements in nuclear proliferation.
• Investment in agile policies and tools is crucial for international organizations and governments to manage evolving nuclear risks.