Reproducibility in machine-learning-based research: Overview, barriers, and drivers
Harald Semmelrock, Tony Ross-Hellauer, Simone Kopeinik, Dieter Theiler, Armin Haberl, Stefan Thalmann, Dominik Kowald
AI Magazine, Volume 46, Issue 2, 2025
Abstract
Many research fields are currently reckoning with issues of poor levels of reproducibility. Some label it a "crisis," and research employing or building machine learning (ML) models is no exception. Issues including lack of transparency, data or code, poor adherence to standards, and the sensitivity of ML training conditions mean that many papers are not even reproducible in principle. Where they are, though, reproducibility experiments have found worryingly low degrees of similarity with original results. Despite previous appeals from ML researchers on this topic and various initiatives from conference reproducibility tracks to the ACM's new Emerging Interest Group on Reproducibility and Replicability, we contend that the general community continues to take this issue too lightly. Poor reproducibility threatens trust in and integrity of research results. Therefore, in this article, we lay out a new perspective on the key barriers and drivers (both procedural and technical) to increased reproducibility at various levels (methods, code, data, and experiments). We then map the drivers to the barriers to give concrete advice for strategies for researchers to mitigate reproducibility issues in their own work, to lay out key areas where further research is needed in specific areas, and to further ignite discussion on the threat presented by these urgent issues.
Tax Administration AI: The Holy Grail to Overcome Information Asymmetry in Tax Enforcement?
Tina Ehrke-Rabel, Barbara Gunacker-Slawitsch
Intertax
Volume 53, Issue 2 (2025) pp. 128 - 140
Abstract
Liberal democracies are based on the concept of the free person who is a rational being capable of understanding their behaviour and taking responsibility for it. The state may only interfere with individual freedom when it has a legal basis, if it is necessary to safeguard the functioning of a democratic society. Therefore, in liberal democracies, citizens have duties to fulfil and responsibilities to bear. In return, they are not and must not be surveilled by the state. Tax enforcement is built upon this concept.
Tax administration Artificial Intelligence (AI) has the potential to change this concept. What is technically feasible is not necessarily appropriate for sustaining the fundamental values of a liberal society.
If society wants to uphold these values, any deployment of tax administration AI must be legally framed with clear rules on the origin of the data, the method of data processing, the aim of the data processing, and the possibility of human intervention. Additional legal prerequisites may be necessary depending on the purpose of the use of tax administration AI. If it is deployed in a mindful way, it is likely to increase both the efficiency and the equality of tax enforcement.
