Who is Responsible for Adversarial Defense?

TL;DR

This paper discusses the importance of identifying responsible parties and expertise needed for implementing and validating adversarial defenses in AI/ML systems, emphasizing accountability and evaluation.

Contribution

It highlights the need for clear responsibility and expertise in developing, validating, and deploying adversarial defenses, fostering a discussion on accountability in AI security.

Findings

Identifies key components vulnerable to adversarial attacks

Emphasizes the importance of expertise in defense validation

Calls for clear responsibility in adversarial defense efforts

Abstract

We have seen a surge in research aims toward adversarial attacks and defenses in AI/ML systems. While it is crucial to formulate new attack methods and devise novel defense strategies for robustness, it is also imperative to recognize who is responsible for implementing, validating, and justifying the necessity of these defenses. In particular, which components of the system are vulnerable to what type of adversarial attacks, and the expertise needed to realize the severity of adversarial attacks. Also how to evaluate and address the adversarial challenges in order to recommend defense strategies for different applications. This paper opened a discussion on who should examine and implement the adversarial defenses and the reason behind such efforts.