Addressing the Gap in International Criminal Liability through Layered Accountability and Algoritmic Sanctions for Autonomous AI

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Background: Autonomous AI poses a challenge to international criminal law because liability based on human actus reus and mens rea is no longer sufficient to address crimes involving AI. Objective: This study aims to analyze and reconstruct the framework of international criminal accountability for crimes involving autonomous AI by examining the principles of ethics, accountability, and AI regulation in international and regional instruments, and their implications for attribution and criminal liability under international criminal law. Methodology: This study employs a normative legal research approach, incorporating conceptual, statutory, comparative, and international-law perspectives to analyze the framework governing criminal liability for international crimes involving autonomous AI. Findings: The analysis shows that accountability for international crimes involving autonomous AI requires developing the concept of corporate liability, attributing responsibility to the relevant human actors, and imposing unconventional sanctions to prevent and control AI behavior. Originality/Uniqueness: This study proposes a risk-based, layered accountability model to address the responsibility gap in international crimes involving autonomous AI by attributing responsibility to human actors and applying unconventional sanctions.