Learning with Conflicts of Interest

TL;DR

This paper introduces a game-theoretic framework to address conflicts of interest in machine learning systems, aiming to protect users from biased information while preserving system utility.

Contribution

It proposes a novel game-theoretic model and scalable algorithms with theoretical guarantees to balance information maximization and bias minimization.

Findings

Algorithms effectively reduce biased and manipulative actions.

The framework maximizes desired information while safeguarding user interests.

The approach is scalable and theoretically grounded.

Abstract

Financial, social, and political factors often prevent the interests of the owners of ML systems and services and their users from being perfectly aligned. ML systems often produce biased information that can influence users to make decisions that are not in their best interest. Current solution approaches require ML systems to implement protocols to mitigate their biases. However, ML system owners usually do not have any incentive to implement these protocols and often argue that it limits their freedom of expression or business. We believe that a successful solution to this problem must recognize the conflict of interest between the ML systems and their users, and use this information to protect users against information that adversely influences their decisions while allowing users to safely benefit from these systems. To this end, we propose a game-theoretic framework that models the interaction between ML systems and users with conflicts of interest. We present scalable algorithms with theoretical guarantees that maximize the amount of desired information and actions and minimize the amount of biased and manipulative actions in interaction with ML systems.