Remote explainability faces the bouncer problem - LAAS - Laboratoire d'Analyse et d'Architecture des Systèmes Accéder directement au contenu
Article Dans Une Revue Nature Machine Intelligence Année : 2020

Remote explainability faces the bouncer problem

Résumé

The concept of explainability is envisioned to satisfy society’s demands for transparency about machine learning decisions. The concept is simple: like humans, algorithms should explain the rationale behind their decisions so that their fairness can be assessed. Although this approach is promising in a local context (for example, the model creator explains it during debugging at the time of training), we argue that this reasoning cannot simply be transposed to a remote context, where a model trained by a service provider is only accessible to a user through a network and its application programming interface. This is problematic, as it constitutes precisely the target use case requiring transparency from a societal perspective. Through an analogy with a club bouncer (who may provide untruthful explanations upon customer rejection), we show that providing explanations cannot prevent a remote service from lying about the true reasons leading to its decisions. More precisely, we observe the impossibility of remote explainability for single explanations by constructing an attack on explanations that hides discriminatory features from the querying user. We provide an example implementation of this attack. We then show that the probability that an observer spots the attack, using several explanations for attempting to find incoherences, is low in practical settings. This undermines the very concept of remote explainability in general.
Fichier principal
Vignette du fichier
1910.01432.pdf (540.42 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03048809 , version 1 (31-12-2020)

Identifiants

Citer

Erwan Le Merrer, Gilles Trédan. Remote explainability faces the bouncer problem. Nature Machine Intelligence, 2020, 2 (9), pp.529-539. ⟨10.1038/s42256-020-0216-z⟩. ⟨hal-03048809⟩
128 Consultations
19 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More