Legal Protection for Consumers Against the Risk of Algorithmic Inaccuracy in Artificial Intelligence Subscription Services

Abstract

Introduction: This research is prompted by the increasing use of subscription-based artificial intelligence services in Indonesia, which poses risks of algorithmic inaccuracy for consumers. The existing legal framework, particularly the Consumer Protection act, the Electronic Information and Transactions act, and Government Regulation on the Implementation of Electronic Systems and Transactions, has not adequately regulated the liability of business actors for substantive errors in the outputs of artificial intelligence systems that harm consumers.

Purposes of the Research: This research aims to identify normative gaps within the Indonesian consumer protection legal framework concerning algorithmic inaccuracy in AI subscription services, and to formulate legal reform proposals oriented toward substantive consumer protection through the reconstruction of legal liability and regulatory reforms that are adaptive to technological developments.

Methods of the Research: This research employs a normative juridical method using both statutory and conceptual approaches.

Findings of the Research: The findings indicate the existence of a three-layered normative gap: (1) a substantive gap regarding the standard of accuracy of AI outputs within Consumer Protection act and Government Regulation on the Implementation of Electronic Systems and Transactions; (2) an imbalance of standard clauses in subscription contracts that shift risks onto consumers; and (3) limitations in the institutional capacity of supervisory bodies. As a novel contribution, this research proposes the doctrine of Algorithmic Duty of Care as a standard of liability for business actors that cannot be waived through contractual clauses.