Private Machine Learning in TensorFlow using Secure Computation

TL;DR

This paper introduces a framework that integrates secure multi-party computation into TensorFlow, enabling privacy-preserving machine learning with existing tools and optimizations, demonstrated through benchmarks on typical models.

Contribution

It provides an open source implementation of a state-of-the-art secure computation protocol within TensorFlow, enhancing privacy-preserving ML workflows.

Findings

Successful integration of secure computation in TensorFlow

Improved performance benchmarks on private ML models

Enhanced tooling for privacy-preserving machine learning

Abstract

We present a framework for experimenting with secure multi-party computation directly in TensorFlow. By doing so we benefit from several properties valuable to both researchers and practitioners, including tight integration with ordinary machine learning processes, existing optimizations for distributed computation in TensorFlow, high-level abstractions for expressing complex algorithms and protocols, and an expanded set of familiar tooling. We give an open source implementation of a state-of-the-art protocol and report on concrete benchmarks using typical models from private machine learning.