The Use of Artificial Intelligence by Tax Authorities for Tax Auditing Purposes: Reflections in Light of the European Taxpayer Code and the Adoption of a Tax Control Framework by Corporate Taxpayers

The use of AI systems for tax audit selection purposes to target both individual and corporate taxpayers raises issues in light of the European Taxpayer Code, particularly regarding the assumption of taxpayer honesty. Moreover, for corporate taxpayers adopting a Tax Control Framework (TCF) aimed at minimizing the tax risk, it requires – due to the TCF’s very purpose – consistency between the outcomes of AI application and the TCF’s implementation. Given the different scope of the use of AI systems among EU Member States, it raises the issue as to whether, in AI use, there is scope for a “procedural competition” – or for a coordination – between Member States. Responses, also those based on the experiences of some Member States, are proposed for these issues. Firstly, by suggesting a proportionality criterion for assessing the legitimacy of the use of AI systems for tax audit selection purposes and secondly – in the case of corporate taxpayers – by arguing that this criterion is met when the outcome of the application of AI systems is consistent with an effective implementation of a TCF.