From Counterfactuals to Trees: Competitive Analysis of Model Extraction Attacks
This addresses security vulnerabilities for MLaaS providers by analyzing and mitigating model extraction risks, though it is incremental as it builds on existing attack frameworks.
The paper tackles the security risks of model extraction attacks in Machine Learning as a Service, where explainability techniques like counterfactual explanations can enable unauthorized model replication, and it presents a formal analysis and reconstruction algorithms for tree-based models that achieve provably perfect fidelity with theoretical bounds on query complexity.
The advent of Machine Learning as a Service (MLaaS) has heightened the trade-off between model explainability and security. In particular, explainability techniques, such as counterfactual explanations, inadvertently increase the risk of model extraction attacks, enabling unauthorized replication of proprietary models. In this paper, we formalize and characterize the risks and inherent complexity of model reconstruction, focusing on the "oracle'' queries required for faithfully inferring the underlying prediction function. We present the first formal analysis of model extraction attacks through the lens of competitive analysis, establishing a foundational framework to evaluate their efficiency. Focusing on models based on additive decision trees (e.g., decision trees, gradient boosting, and random forests), we introduce novel reconstruction algorithms that achieve provably perfect fidelity while demonstrating strong anytime performance. Our framework provides theoretical bounds on the query complexity for extracting tree-based model, offering new insights into the security vulnerabilities of their deployment.