THE LEGAL STATUS OF AUTONOMOUS CYBER AGENTS: RETHINKING CRIMINAL LIABILITY IN THE AGE OF AI.

Abstract

As artificial intelligence (AI) systems evolve from passive tools to autonomous cyber agents capable of making decisions and executing tasks without direct human oversight, legal systems worldwide face a profound challenge: how to assign criminal liability when such agents commit unlawful acts. This paper critically examines the legal status of autonomous cyber agents and explores the doctrinal and jurisprudential implications of attributing criminal responsibility in the age of AI. Through a comparative analysis of the legal frameworks in the United Kingdom, Nigeria, and the State of Illinois (USA), the study interrogates whether existing criminal law doctrines such as mens rea, actus reus, and corporate liability are sufficient to address offenses committed by or through intelligent systems. The paper evaluates the extent to which these jurisdictions recognize or resist the notion of non-human agency in criminal law, and whether emerging legal theories such as electronic personhood, algorithmic accountability, or strict liability offer viable pathways for reform. The research highlights key gaps in statutory interpretation, evidentiary standards, and prosecutorial discretion when dealing with AI-driven cybercrime. It also considers the ethical and policy dimensions of imposing liability on developers, users, or the autonomous systems themselves. Ultimately, the paper proposes a hybrid regulatory model that balances innovation with accountability, offering recommendations for legislative and judicial adaptation in light of rapidly advancing AI technologies.