Yogyakarta — The use of artificial intelligence (AI) in recruitment can improve hiring efficiency, but it may also carry gender bias into employment decisions. This finding comes from research by Nur Aeinun Nisya and Rusmita Ayu Rahmawati from the Faculty of Economics and Business, Yogyakarta State University. The study mapped academic research on AI, gender bias, recruitment, and workplace inclusion by analyzing 92 Scopus-indexed journal articles published between 2016 and 2025. The findings show that AI-based recruitment should not be viewed only as a tool for speeding up hiring, but also as a system closely connected to fairness, transparency, and inclusion in the workplace.
Artificial intelligence is increasingly being used in human resource management, including screening resumes, matching candidates with job positions, conducting preliminary assessments, and supporting hiring decisions. These technologies allow organizations to process large amounts of applicant data more quickly and consistently than traditional recruitment methods.
However, greater technological efficiency does not necessarily mean that recruitment becomes fairer. AI systems rely on data and algorithms used during the training process. If those datasets contain patterns of inequality from previous recruitment practices, AI systems may learn and reproduce the same patterns. As a result, technology that appears objective may instead maintain or even amplify existing discrimination.
Gender bias has become one of the major concerns surrounding the use of AI in recruitment. When systems are trained using historical data that reflect gender inequalities, candidates from certain groups may receive different treatment even when they have comparable qualifications. Bias can also emerge from language patterns, data imbalance, and algorithm design choices.
The issue is particularly important because recruitment serves as the main gateway to employment. Decisions made during resume screening, interview selection, and final hiring can determine who receives access to job opportunities. If biased algorithmic recommendations influence human decision-makers, inequality can extend into the workplace.
To examine these developments, Nisya and Rahmawati conducted a bibliometric analysis of academic publications. The research data were retrieved from the Scopus database on June 4, 2025. The search was limited to English-language journal articles published between 2016 and 2025 that addressed AI or algorithmic systems in recruitment, bias or discrimination, and inclusion, diversity, or equality in organizational settings. After applying the selection criteria and removing duplicates, 92 relevant articles were included in the analysis.
The researchers used VOSviewer to examine relationships among keywords in the selected publications. The analysis showed that “recruitment” was the most frequently occurring keyword, appearing 54 times and recording a total link strength of 76. This position indicates that recruitment has become the central topic connecting AI, bias, fairness, and inclusion in the academic literature.
Other strongly connected keywords included “artificial intelligence,” “discrimination,” and “human resource management.” The findings indicate a shift in how researchers view AI in human resource management. AI is no longer discussed only as a technology for improving speed and efficiency, but increasingly as a factor that can influence equity, diversity, and workplace inclusion.
The analysis identified three major thematic clusters. The first focuses on social and ethical issues, including recruitment, bias, discrimination, gender, and selection. The second covers technology and human resource management, including artificial intelligence, machine learning, and recruitment processes. The third focuses on decision-making and inclusion, involving decision-making, hiring, employee selection, gender bias, and inclusion.
These findings indicate that gender bias in AI-based recruitment is not simply a technical problem. Bias is connected to how data are collected, how algorithms are designed, how technological recommendations are used, and how human decision-makers respond to AI-generated results.
The research also identified an evolution in academic attention. Earlier studies focused more heavily on gender, bias, discrimination, and unequal access to employment. More recent research increasingly emphasizes artificial intelligence, diversity, and inclusion. This shift reflects a move from simply identifying discrimination toward developing fairer and more inclusive recruitment systems.
However, the study found that inclusion remains less developed in the literature than bias and discrimination. The VOSviewer density visualization showed that “recruitment,” “artificial intelligence,” “bias,” “gender,” “discrimination,” and “human resource management” remain the core research topics. Meanwhile, “diversity” and “inclusion” are emerging themes but have not yet reached the same level of research concentration.
This situation reveals an important research gap. Many studies continue to focus on identifying bias, while fewer examine how organizations can actively design AI-based recruitment systems that promote inclusion. Strategies such as algorithmic audits, fairness-by-design approaches, transparent training data, and inclusive HR governance require greater research attention.
For businesses, the findings suggest that AI recruitment systems should not be evaluated solely based on speed, efficiency, or predictive performance. Organizations also need to ensure that automated systems do not reinforce inequalities contained in historical data. Transparency, algorithmic fairness assessments, system audits, and human oversight are important elements of responsible AI implementation.
For policymakers and human resource practitioners, the study highlights the need to view AI as part of a broader social and organizational system rather than simply as a technological tool. Human judgment remains important to ensure that algorithmic recommendations do not automatically determine an applicant’s employment opportunities.
Nisya and Rahmawati conclude that the future of AI-based recruitment should not be measured only by technological sophistication. Its success should also be assessed by its ability to create recruitment processes that are fairer, more transparent, and genuinely inclusive for candidates from diverse backgrounds.
The study acknowledges several limitations. The analysis was restricted to publications indexed in Scopus and relied on keyword co-occurrence analysis, which may not capture every conceptual aspect of the field. Future research could expand the database sources to include platforms such as Web of Science and Google Scholar and use citation, co-authorship, or cross-country and cross-industry analysis to provide broader insights into AI fairness in recruitment.
Author Profiles
Nur Aeinun Nisya — Faculty of Economics and Business, Yogyakarta State University.
Rusmita Ayu Rahmawati — Faculty of Economics and Business, Yogyakarta State University.
Research Source
Article Title: Mapping Gender Bias and Workplace Inclusion in Artificial Intelligence-Based Recruitment: A Bibliometric Co-Occurrence
Journal: East Asian Journal of Multidisciplinary Research (EAJMR), Vol. 5 No. 8, 2026, pp. 3345–3362.
DOI: https://doi.org/10.55927/eajmr.v5i8.264
Journal Website: https://journaleajmr.my.id/index.php/eajmr
0 Komentar