MLCapsule: Guarded Offline Deployment of Machine Learning as a Service

TL;DR

MLCapsule enables secure, offline deployment of machine learning models on client devices, protecting user data and model intellectual property while defending against advanced attacks like model stealing and membership inference.

Contribution

The paper introduces MLCapsule, a novel system for guarded offline ML deployment that maintains model security and privacy without relying on server-side execution.

Findings

MLCapsule effectively protects user data during offline ML model execution.

It prevents model theft, reverse engineering, and membership inference attacks.

The system maintains comparable security levels to server-based ML services.

Abstract

With the widespread use of machine learning (ML) techniques, ML as a service has become increasingly popular. In this setting, an ML model resides on a server and users can query it with their data via an API. However, if the user's input is sensitive, sending it to the server is undesirable and sometimes even legally not possible. Equally, the service provider does not want to share the model by sending it to the client for protecting its intellectual property and pay-per-query business model. In this paper, we propose MLCapsule, a guarded offline deployment of machine learning as a service. MLCapsule executes the model locally on the user's side and therefore the data never leaves the client. Meanwhile, MLCapsule offers the service provider the same level of control and security of its model as the commonly used server-side execution. In addition, MLCapsule is applicable to offline applications that require local execution. Beyond protecting against direct model access, we couple the secure offline deployment with defenses against advanced attacks on machine learning models such as model stealing, reverse engineering, and membership inference.