Explainable AI in Taxation: Melanie Fink on the Role of Explanation Requirements in EU Constitutional Law
On 10 March 2023, the Amsterdam Centre for Tax Law (ACTL) of the University of Amsterdam (UvA) organised the Conference ‘Towards eXplainable Artificial Intelligence (XAI) in Taxation: The Future of Good Tax Governanceon’.
The conference brought together 18 speakers and moderators from academia, tax administration, and technology to discuss ways to overcome legal and technical obstacles to ensure eXplainability of AI systems used for tax purposes. The goal was to explore the relationship between minimizing the potential harm to taxpayers and maximizing the efficiency of tax authorities and the role that following XAI guidelines in the design, deployment, and implementation of tax AI may play in that respect.
Melanie Fink delivered a presentation on ‘Reasoned A(I)dministration: Explanation Requirements in EU Law and the Automation of Public Administration’, based on a paper co-authored with Michèle Finck, Professor of Law and Artificial Intelligence at the University of Tübingen. Their paper looks at the role of the duty to give reasons to meaningfully control the EU administration when it relies on AI systems to inform their decision-making and the added value of secondary law, in particular data protection law and the draft EU Artificial Intelligence Act, in this respect. They argue that the duty to give reasons provides a useful starting point, but leaves a number of aspects unclear. Neither EU data protection law nor the AIA currently fill these gaps. The paper is available for download here.
The conference was organized by the Amsterdam Centre for Tax Law (ACTL) of the University of Amsterdam (UvA) under the umbrella of the research project ‘Designing the tax system for a Cashless, Platform-based and Technology-driven society’ (CPT project) and was supported in part by The Notre Dame-IBM Tech Ethics Lab’s scientific grant.