Using predictive multiplicity to measure individual performance within the AI Act
This work tackles the challenge of ensuring fairness and reliability in high-risk AI systems under EU regulations, though it is incremental as it builds on existing predictive multiplicity concepts to address legal compliance.
This paper addresses the problem of predictive multiplicity in AI systems, where multiple models with similar overall accuracy make conflicting predictions for individual cases, arguing that this arbitrariness conflicts with the EU AI Act's requirement to report performance for specific persons. The result is a framework suggesting tools like individual conflict ratios and δ-ambiguity to quantify and report this multiplicity for compliance.
When building AI systems for decision support, one often encounters the phenomenon of predictive multiplicity: a single best model does not exist; instead, one can construct many models with similar overall accuracy that differ in their predictions for individual cases. Especially when decisions have a direct impact on humans, this can be highly unsatisfactory. For a person subject to high disagreement between models, one could as well have chosen a different model of similar overall accuracy that would have decided the person's case differently. We argue that this arbitrariness conflicts with the EU AI Act, which requires providers of high-risk AI systems to report performance not only at the dataset level but also for specific persons. The goal of this paper is to put predictive multiplicity in context with the EU AI Act's provisions on accuracy and to subsequently derive concrete suggestions on how to evaluate and report predictive multiplicity in practice. Specifically: (1) We argue that incorporating information about predictive multiplicity can serve compliance with the EU AI Act's accuracy provisions for providers. (2) Based on this legal analysis, we suggest individual conflict ratios and $δ$-ambiguity as tools to quantify the disagreement between models on individual cases and to help detect individuals subject to conflicting predictions. (3) Based on computational insights, we derive easy-to-implement rules on how model providers could evaluate predictive multiplicity in practice. (4) Ultimately, we suggest that information about predictive multiplicity should be made available to deployers under the AI Act, enabling them to judge whether system outputs for specific individuals are reliable enough for their use case.