Under Which Conditions Would the ATAD’s GAAR Benefit from Automatic Decision-Making in Terms of Legal Certainty and Efficiency?

Abstract

In this article, the author scrutinized the limitations on tax administrations when it comes to enhancing legal certainty and efficiency via automatica decision-making (ADM) models until the Anti-Tax Avoidance Directive and general anti-avoidance rules. A lack of clarity and certainty is identified under the Anti-Tax Avoidance Directive's general anti-avoidance rule framework. To curb this issue, the Court of Justice of the European Union's case lae is a vital source for enhancing integrability with any ADM model. Taxpayers' rights are found to be limited by transparency breaches and data misuse under ADM, mainly in respect of black-box models. An ideal hybrid approach is identified for artificial intelligence, via machine learning and traditional computer mechanisms, since it uses sufficiently less data and requires only a good model understanding. Based on a practical simulation approach, facts and patterns are extracted from case law. If certain conditions are met, findings show that, departing from the Anti-Tax Avoidance Directive's general anti-abuse rule tests, the motive and artificiality test are more prone to automatization than the defeat-of-object-of-purpose test. All of them benefit from the case law of the Court of Justice of the European Union on automated scenarios. Thus, the application of enhanced ADM model issues may be addressed by including human beings in the loop.