26.10.2026
Best Paper Award at AIES 2026 for Work on Algorithmic Contestability
Why Explainable AI Is Not Enough to Challenge Algorithmic Decisions
The paper “Explainable AI Isn’t Enough! Rethinking Algorithmic Contestability” received the Best Paper Award at the Ninth AAAI/ACM Conference on AI, Ethics, and Society (AIES 2026). The paper was written by MCML Junior Member Timo Freiesleben with co-authors Kristof Meding, and Gunnar König.
Machine learning systems increasingly make life-changing decisions about individuals, for example, in loan approvals or hiring. Explainable AI research has so far focused mainly on recourse, which helps individuals change their situation to obtain a better outcome. Much less attention has been paid to contestability, which helps individuals challenge and correct decisions that may be wrong. The authors propose an operational definition of contestability and show that standard explanation methods, such as counterfactuals, LIME, or Anchors, are not sufficient to overturn a decision. Instead, they identify three types of evidence that can justify reversing a decision: predictive multiplicity, incorrect feature values, and neglected evidence that should have overruled the decision. They also analyze how existing EU legislation already grants individuals some rights to such evidence.
Congratulations to the authors on this recognition!
Related
12.11.2026
One Toolbox for Many Kinds of Uncertainty
MCML researchers introduce probly, a unified Python package for uncertainty-aware machine learning, presented orally at NeurIPS 2026.
10.11.2026
Enkelejda Kasneci: The Real Value of AI in Education Is More Time for Humans
Enkelejda Kasneci explores how AI can provide personalized learning support while keeping human interaction and student agency at the center.
09.11.2026
Outstanding Paper Award at EMNLP 2026 for Work on Accent Bias in Audio Language Models
A paper co-authored by MCML PI Valentin Hofmann on accent bias in audio language models won an Outstanding Paper Award at EMNLP 2026.