Developments over the last decade in the use of AI in tax administration have been nothing short of outstanding. Not only are taxpayers increasingly making use of automated systems in tax compliance, but perhaps more importantly, tax enforcement is increasingly reliant on new technologies as compliance-enhancing and fraud-prevention tools. However, whilst the use of AI brings very significant advantages to both the efficiency and the equity of tax systems, it also carries important risks. This paper identifies the development of a new AI fallacy within the tax policy sphere, namely that of unconstrained success: that the use of AI in tax compliance and enforcement can compensate for the deficiencies of the tax law. This paper considers the rationale behind the development of this fallacy, in particular the political and institutional dynamics involved in the approval of new tax legislation. It concludes that, maximising the benefits of the use of AI in tax compliance and enforcement requires departure from this fallacy, and the recognition of the wider dynamics of the tax policy-administration symbiosis.
Source Rita de la Feria
