Identifying Appropriate Intellectual Property Protection Mechanisms for Machine Learning Models: A Systematization of Watermarking, Fingerprinting, Model Access, and Attacks

TL;DR

This paper provides a comprehensive systematization of intellectual property protection mechanisms for machine learning models, including threats, attacks, and defenses, by developing a unified threat model and taxonomy to bridge ML and security research.

Contribution

It introduces a unified threat model and taxonomy for IP protection in ML, consolidating fragmented research and bridging ML and security perspectives.

Findings

Developed a comprehensive threat model for ML IP protection

Categorized attacks and defenses within a unified taxonomy

Bridged research gaps between ML and security communities

Abstract

The commercial use of Machine Learning (ML) is spreading; at the same time, ML models are becoming more complex and more expensive to train, which makes Intellectual Property Protection (IPP) of trained models a pressing issue. Unlike other domains that can build on a solid understanding of the threats, attacks and defenses available to protect their IP, the ML-related research in this regard is still very fragmented. This is also due to a missing unified view as well as a common taxonomy of these aspects. In this paper, we systematize our findings on IPP in ML, while focusing on threats and attacks identified and defenses proposed at the time of writing. We develop a comprehensive threat model for IP in ML, categorizing attacks and defenses within a unified and consolidated taxonomy, thus bridging research from both the ML and security communities.