Regulation Games for Trustworthy Machine Learning
This addresses the challenge of coordinating multiple trust aspects like fairness and privacy in ML for policymakers and developers, though it is incremental as it builds on existing game-theoretic and multi-objective optimization ideas.
The paper tackles the problem of trustworthy machine learning by proposing a game-theoretic framework called regulation games to model interactions between model builders and regulators, showing that in a gender classification application, regulators can enforce a differential privacy budget 4.0 lower on average by specifying guarantees first.
Existing work on trustworthy machine learning (ML) often concentrates on individual aspects of trust, such as fairness or privacy. Additionally, many techniques overlook the distinction between those who train ML models and those responsible for assessing their trustworthiness. To address these issues, we propose a framework that views trustworthy ML as a multi-objective multi-agent optimization problem. This naturally lends itself to a game-theoretic formulation we call regulation games. We illustrate a particular game instance, the SpecGame in which we model the relationship between an ML model builder and fairness and privacy regulators. Regulators wish to design penalties that enforce compliance with their specification, but do not want to discourage builders from participation. Seeking such socially optimal (i.e., efficient for all agents) solutions to the game, we introduce ParetoPlay. This novel equilibrium search algorithm ensures that agents remain on the Pareto frontier of their objectives and avoids the inefficiencies of other equilibria. Simulating SpecGame through ParetoPlay can provide policy guidance for ML Regulation. For instance, we show that for a gender classification application, regulators can enforce a differential privacy budget that is on average 4.0 lower if they take the initiative to specify their desired guarantee first.