TITLE:
AI and Risks of Hiring Bias Due to Gender Imbalances in Historical Data
AUTHORS:
Julia M. Burch, Ross A. Jackson, Claire I. Craig
KEYWORDS:
Analytics, Diversity, Ethics, Feminism, Gender, Inequality, Recruitment
JOURNAL NAME:
Voice of the Publisher,
Vol.12 No.1,
March
13,
2026
ABSTRACT: Unless underlying data are corrected, the use of Artificial Intelligence (AI) in organizational hiring processes holds the potential to perpetuate longstanding gender biases. This study investigated the risk of gender biases resident in a potential data source which could be used to train AI used for organizational recruitment. As AI proliferates and becomes increasingly integrated into organizational hiring practices, it is crucial to understand how historical data, often laden with gender biases, influences AI-system outcomes. This study utilized Google Books Ngram Viewer to analyze the prevalence of gender-coded words and phrases over the past century, the years 1922 to 2022 inclusive, highlighting the dominance of masculine terms and traits. Hypothesis testing revealed that the average percentage of occurrence of masculine pronouns were significantly greater than that of feminine pronouns, when tested using a t-test for two samples assuming unequal variances (α = 0.05). Masculine pronouns were approximately 2.7 times greater than feminine pronouns. Further, there was a greater number of female counts associated with feminine traits, when tested using the chi-square test (α = 0.05). When AFINN sentiment was considered, these disparities resulted in a one-point semantic advantage for males over females, with an average male sentiment value of 0.4 (slightly positive), and an average female sentiment value of −0.4 (slightly negative). Collectively, these findings suggest that AI systems trained on biased data may perpetuate gender inequality by favoring male candidates and reinforcing patriarchal stereotypes, privilege, and power. The paper calls for employing bias detection tools, enhancing transparency, and fostering governance to mitigate these biases. Further, this work emphasized a need for more inclusive data practices and active female participation in AI development to ensure fairness and equity in hiring. This research underscores the importance of critically examining AI’s role in shaping modern workforce dynamics, and is of potential interest to those engaged in AI development, organizational management, and Human Resources (HR).