Who's Gaming the System? A Causally-Motivated Approach for Detecting Strategic Adaptation
This addresses the issue of strategic adaptation in multi-agent systems for stakeholders like policymakers or system designers, but it is incremental as it builds on existing causal methods for a specific detection task.
The paper tackles the problem of detecting which agents are most aggressively gaming machine learning models by manipulating inputs, and shows that a ranking of agents by their gaming tendency is identifiable through causal effect estimation, with empirical validation on synthetic data and a U.S. diagnosis coding case study.
In many settings, machine learning models may be used to inform decisions that impact individuals or entities who interact with the model. Such entities, or agents, may game model decisions by manipulating their inputs to the model to obtain better outcomes and maximize some utility. We consider a multi-agent setting where the goal is to identify the "worst offenders:" agents that are gaming most aggressively. However, identifying such agents is difficult without knowledge of their utility function. Thus, we introduce a framework in which each agent's tendency to game is parameterized via a scalar. We show that this gaming parameter is only partially identifiable. By recasting the problem as a causal effect estimation problem where different agents represent different "treatments," we prove that a ranking of all agents by their gaming parameters is identifiable. We present empirical results in a synthetic data study validating the usage of causal effect estimation for gaming detection and show in a case study of diagnosis coding behavior in the U.S. that our approach highlights features associated with gaming.