Founding The Domain of AI Forensics

TL;DR

This paper advocates for establishing AI Forensics as a new discipline to investigate AI failures scientifically and legally, inspired by cyber forensics, and discusses its foundational challenges.

Contribution

It introduces the concept of AI Forensics, proposes a taxonomy of its subfields, and discusses key foundational challenges for this emerging discipline.

Findings

AI Forensics is needed for legal and scientific investigation of AI failures.

A taxonomy of AI Forensics subfields is proposed.

Foundational challenges for AI Forensics are discussed.

Abstract

With the widespread integration of AI in everyday and critical technologies, it seems inevitable to witness increasing instances of failure in AI systems. In such cases, there arises a need for technical investigations that produce legally acceptable and scientifically indisputable findings and conclusions on the causes of such failures. Inspired by the domain of cyber forensics, this paper introduces the need for the establishment of AI Forensics as a new discipline under AI safety. Furthermore, we propose a taxonomy of the subfields under this discipline, and present a discussion on the foundational challenges that lay ahead of this new research area.