Trustworthiness Preservation by Copies of Machine Learning Systems

TL;DR

This paper introduces a formal calculus to verify if copies of machine learning systems maintain trustworthiness, providing a framework to analyze complex probabilistic queries and trust notions in AI systems.

Contribution

It proposes a novel calculus for modeling and verifying trustworthiness in ML system copies, addressing an underexplored aspect of responsible AI.

Findings

Defined four trustworthiness notions: Justifiably, Equally, Weakly, Almost Trustworthy

Analyzed relations and compositions of trust notions

Provided a computational tool for trust verification in copied systems

Abstract

A common practice of ML systems development concerns the training of the same model under different data sets, and the use of the same (training and test) sets for different learning models. The first case is a desirable practice for identifying high quality and unbiased training conditions. The latter case coincides with the search for optimal models under a common dataset for training. These differently obtained systems have been considered akin to copies. In the quest for responsible AI, a legitimate but hardly investigated question is how to verify that trustworthiness is preserved by copies. In this paper we introduce a calculus to model and verify probabilistic complex queries over data and define four distinct notions: Justifiably, Equally, Weakly and Almost Trustworthy which can be checked analysing the (partial) behaviour of the copy with respect to its original. We provide a study of the relations between these notions of trustworthiness, and how they compose with each other and under logical operations. The aim is to offer a computational tool to check the trustworthiness of possibly complex systems copied from an original whose behavour is known.