Measuring What Matters: Connecting AI Ethics Evaluations to System Attributes, Hazards, and Harms

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Rismani, Shalaleh, Shelby, Renee, Davis, Leah, Rostamzadeh, Negar, Moon, AJung
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909839631843328
author Rismani, Shalaleh
Shelby, Renee
Davis, Leah
Rostamzadeh, Negar
Moon, AJung
author_facet Rismani, Shalaleh
Shelby, Renee
Davis, Leah
Rostamzadeh, Negar
Moon, AJung
contents Over the past decade, an ecosystem of measures has emerged to evaluate the social and ethical implications of AI systems, largely shaped by high-level ethics principles. These measures are developed and used in fragmented ways, without adequate attention to how they are situated in AI systems. In this paper, we examine how existing measures used in the computing literature map to AI system components, attributes, hazards, and harms. Our analysis draws on a scoping review resulting in nearly 800 measures corresponding to 11 AI ethics principles. We find that most measures focus on four principles - fairness, transparency, privacy, and trust - and primarily assess model or output system components. Few measures account for interactions across system elements, and only a narrow set of hazards is typically considered for each harm type. Many measures are disconnected from where harm is experienced and lack guidance for setting meaningful thresholds. These patterns reveal how current evaluation practices remain fragmented, measuring in pieces rather than capturing how harms emerge across systems. Framing measures with respect to system attributes, hazards, and harms can strengthen regulatory oversight, support actionable practices in industry, and ground future research in systems-level understanding.
format Preprint
id arxiv_https___arxiv_org_abs_2510_10339
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Measuring What Matters: Connecting AI Ethics Evaluations to System Attributes, Hazards, and Harms
Rismani, Shalaleh
Shelby, Renee
Davis, Leah
Rostamzadeh, Negar
Moon, AJung
Human-Computer Interaction
Artificial Intelligence
Machine Learning
Over the past decade, an ecosystem of measures has emerged to evaluate the social and ethical implications of AI systems, largely shaped by high-level ethics principles. These measures are developed and used in fragmented ways, without adequate attention to how they are situated in AI systems. In this paper, we examine how existing measures used in the computing literature map to AI system components, attributes, hazards, and harms. Our analysis draws on a scoping review resulting in nearly 800 measures corresponding to 11 AI ethics principles. We find that most measures focus on four principles - fairness, transparency, privacy, and trust - and primarily assess model or output system components. Few measures account for interactions across system elements, and only a narrow set of hazards is typically considered for each harm type. Many measures are disconnected from where harm is experienced and lack guidance for setting meaningful thresholds. These patterns reveal how current evaluation practices remain fragmented, measuring in pieces rather than capturing how harms emerge across systems. Framing measures with respect to system attributes, hazards, and harms can strengthen regulatory oversight, support actionable practices in industry, and ground future research in systems-level understanding.
title Measuring What Matters: Connecting AI Ethics Evaluations to System Attributes, Hazards, and Harms
topic Human-Computer Interaction
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2510.10339