<![CDATA[The rapid diffusion of Artificial Intelligence (AI) has outpaced the capacity of Indonesian civil law to allocate responsibility when AI-driven systems cause loss, leaving victims of deepfake fraud, algorithmic misconduct, and automated decision-making without a settled legal remedy. Existing scholarship has extensively debated whether AI may be recognised as a legal subject, yet has rarely translated that debate into a bounded, court-usable liability framework anchored in the Indonesian Civil Code (KUH Perdata) and tested against comparative risk-based regimes such as the European Union's Artificial Intelligence Act. This article addresses that gap by analysing, through a normative-juridical method with statutory and comparative-conceptual approaches, how civil liability for AI-caused losses should be constructed under Indonesian law. The findings show that AI remains an intangible object, not a legal subject, so liability is properly attributed to the creator, owner, or user of the AI system through the unlawful-act doctrine of Articles 1365 and 1367 of the KUH Perdata, applied together with the data-protection obligations of Government Regulation 71/2019. The article's novelty lies in integrating quantified domestic harm data with a three-track (developer-owner-user) liability model calibrated against the EU AI Act's risk tiers, from which it derives specific regulatory and adjudicative recommendations.]]>
Copyrights © 2026
