The Fallacy of AI Functionality

TL;DR

This paper highlights the importance of assessing AI system functionality, revealing that many deployed AI systems fail to work properly, which is often overlooked in policy discussions focused on ethics.

Contribution

It introduces a taxonomy of AI functionality issues through case studies and emphasizes the need to prioritize functionality assessment in policy and organizational responses.

Findings

Many AI systems fail to function as intended

Current policies often overlook functionality issues

Addressing functionality is crucial for effective AI regulation

Abstract

Deployed AI systems often do not work. They can be constructed haphazardly, deployed indiscriminately, and promoted deceptively. However, despite this reality, scholars, the press, and policymakers pay too little attention to functionality. This leads to technical and policy solutions focused on "ethical" or value-aligned deployments, often skipping over the prior question of whether a given system functions, or provides any benefits at all. To describe the harms of various types of functionality failures, we analyze a set of case studies to create a taxonomy of known AI functionality issues. We then point to policy and organizational responses that are often overlooked and become more readily available once functionality is drawn into focus. We argue that functionality is a meaningful AI policy challenge, operating as a necessary first step towards protecting affected communities from algorithmic harm.