Impacts of Racial Bias in Historical Training Data for News AI

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bhargava, Rahul, Jespersen, Malene Hornstrup, Ndulue, Emily Boardman, Dsouza, Vivica
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918255462973440
author Bhargava, Rahul
Jespersen, Malene Hornstrup
Ndulue, Emily Boardman
Dsouza, Vivica
author_facet Bhargava, Rahul
Jespersen, Malene Hornstrup
Ndulue, Emily Boardman
Dsouza, Vivica
contents AI technologies have rapidly moved into business and research applications that involve large text corpora, including computational journalism research and newsroom settings. These models, trained on extant data from various sources, can be conceptualized as historical artifacts that encode decades-old attitudes and stereotypes. This paper investigates one such example trained on the broadly-used New York Times Annotated Corpus to create a multi-label classifier. Our use in research settings surfaced the concerning "blacks" thematic topic label. Through quantitative and qualitative means we investigate this label's use in the training corpus, what concepts it might be encoding in the trained classifier, and how those concepts impact our model use. Via the application of explainable AI methods, we find that the "blacks" label operates partially as a general "racism detector" across some minoritized groups. However, it performs poorly against expectations on modern examples such as COVID-19 era anti-Asian hate stories, and reporting on the Black Lives Matter movement. This case study of interrogating embedded biases in a model reveals how similar applications in newsroom settings can lead to unexpected outputs that could impact a wide variety of potential uses of any large language model-story discovery, audience targeting, summarization, etc. The fundamental tension this exposes for newsrooms is how to adopt AI-enabled workflow tools while reducing the risk of reproducing historical biases in news coverage.
format Preprint
id arxiv_https___arxiv_org_abs_2512_16901
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Impacts of Racial Bias in Historical Training Data for News AI
Bhargava, Rahul
Jespersen, Malene Hornstrup
Ndulue, Emily Boardman
Dsouza, Vivica
Machine Learning
Artificial Intelligence
Computation and Language
Computers and Society
AI technologies have rapidly moved into business and research applications that involve large text corpora, including computational journalism research and newsroom settings. These models, trained on extant data from various sources, can be conceptualized as historical artifacts that encode decades-old attitudes and stereotypes. This paper investigates one such example trained on the broadly-used New York Times Annotated Corpus to create a multi-label classifier. Our use in research settings surfaced the concerning "blacks" thematic topic label. Through quantitative and qualitative means we investigate this label's use in the training corpus, what concepts it might be encoding in the trained classifier, and how those concepts impact our model use. Via the application of explainable AI methods, we find that the "blacks" label operates partially as a general "racism detector" across some minoritized groups. However, it performs poorly against expectations on modern examples such as COVID-19 era anti-Asian hate stories, and reporting on the Black Lives Matter movement. This case study of interrogating embedded biases in a model reveals how similar applications in newsroom settings can lead to unexpected outputs that could impact a wide variety of potential uses of any large language model-story discovery, audience targeting, summarization, etc. The fundamental tension this exposes for newsrooms is how to adopt AI-enabled workflow tools while reducing the risk of reproducing historical biases in news coverage.
title Impacts of Racial Bias in Historical Training Data for News AI
topic Machine Learning
Artificial Intelligence
Computation and Language
Computers and Society
url https://arxiv.org/abs/2512.16901