RECONSTRUCTING CRIMINAL LIABILITY FOR AI-ENABLED CYBERCRIME IN INDONESIA: Toward a Normative Framework of Responsibility

Abstract

The rapid advancement of artificial intelligence (AI) has fundamentally transformed the nature of cybercrime from human-controlled conduct into algorithmically mediated and semi-autonomous operations. In Indonesia, this transformation is reflected in the sharp increase in cybercrime cases, automated fraud, and AI-enabled digital manipulation, which challenge the traditional criminal liability framework based on human intent and direct control. This study employs normative legal research integrating statutory, conceptual, and comparative approaches to examine the adequacy of existing legal frameworks and to formulate a reconstructed model of criminal liability for AI-based cybercrime. The findings reveal that Indonesia’s current legal regime, primarily governed by the Criminal Code and Electronic Information and Transactions Law, remains rooted in an actor-based liability paradigm, resulting in an accountability gap when harm is caused through autonomous or semi-autonomous AI systems. This article offers a novel normative reconstruction based on a risk- and control-based liability framework operationalized through six elements: recognition of systemic AI risk, expansion of fault doctrine to include algorithmic oversight failure, adoption of multi-tier criminal liability, mandatory algorithmic transparency and audit trails, corporate liability based on risk management failure, and integration of the duty of care principle. This model preserves the principle of personal criminal responsibility while extending liability to actors who control and manage technological risk. The study contributes theoretically by bridging classical fault-based liability and risk-based accountability in criminal law and provides normative recommendations for harmonizing national criminal law and developing AI-specific regulations to ensure legal certainty and effective accountability in the digital era