The Smart Tax System as a Mechanism for Curbing Tax Evasion (A Proposed Artificial Intelligence–Driven Tax Model and a Field Study of Tax Experts in Algeria)

Abstract

This study advances a practical proposal for an intelligent Algerian tax system grounded in e-invoicing, advanced analytics, and artificial intelligence to enhance revenue collection, reduce audit costs, and combat tax evasion. The model centers on an integrated national platform linking the General Directorate of Taxes (DGI) with key stakeholders, including banks and payment institutions, customs authorities, the National Commercial Registry, social security institutions, land registry services, and oversight bodies, while ensuring data protection and algorithmic governance safeguards.Adopting a descriptive-analytical approach complemented by a field survey of 55 tax experts—most with over 25 years of experience—the findings (based on a five-point Likert scale) reveal a low evaluation of the current tax system (≈ 2.55/5), compared to a high assessment of AI integration in tax administration (≈ 4.12/5) and a very high evaluation of the proposed intelligent e-invoicing platform (≈ 4.22/5). International experiences (France, the United Kingdom, the United States, Australia, India, and Brazil) indicate that integrating multi-source data with risk-based models improves audit targeting and strengthens voluntary compliance. The study concludes with phased recommendations adapted to the Algerian context and underscores the model’s adaptability to other jurisdictions, subject to institutional alignment.