Applying XAI to an AI-based system for candidate management to mitigate bias and discrimination in hiring

Abstract

Assuming that potential biases of Artificial Intelligence (AI)-based systems can be identified and controlled for (e.g., by providing high quality training data), employing such systems to augment human resource (HR)-decision makers in candidate selection provides an opportunity to make selection processes more objective. However, as the final hiring decision is likely to remain with humans, prevalent human biases could still cause discrimination. This work investigates the impact of an AI-based system’s candidate recommendations on humans’ hiring decisions and how this relation could be moderated by an Explainable AI (XAI) approach. We used a self-developed platform and conducted an online experiment with 194 participants. Our quantitative and qualitative findings suggest that the recommendations of an AI-based system can reduce discrimination against older and female candidates but appear to cause fewer selections of foreign-race candidates. Contrary to our expectations, the same XAI approach moderated these effects differently depending on the context.

Mapping the landscape of ethical considerations in explainable AI research

“…Papers classified under categories C, D, and E demonstrated varying, yet more substantiated, levels of engagement with ethical theories and frameworks. A closer examination of the papers in each category reveals distinct patterns in the depth and quality of ethical engagement: Computer Science (Hofeditz et al, 2022), (Baum et al, 2022), (El-Nasr & Kleinman, 2020), (Nicodeme, 2020), (Gerdes, 2021), (Dexe et al, 2020), (Lindner & Möllney, 2019), (Falomir & Costa, 2021), (van Otterlo & Atzmueller, 2020) Ethics & Society (Fleisher, 2022), (Larsson & Heintz, 2020), (Narayanan & Tan, 2023), (Kasirzadeh & Smart, 2021) Medicine/Healthcare (Martinho et al, 2021), (Morris et al, 2023), (Jongepier & Keymolen, 2022), (Kempt et al, 2022), (Heinrichs & Eickhoff, 2020), (Herzog, 2022a) 2020) conducts an ethical assessment of explainability in AI-based clinical decision support using the "Principles of Biomedical Ethics," while (Sullivan & Verreault-Julien, 2022) proposes using the capability approach to provide ethical standards for algorithmic recourse. Notably, papers from philosophical, social science, and interdisciplinary backgrounds (e.g., (Amann et al, 2020)) often provide more extensive engagement with ethical theories and frameworks compared to papers from purely technical domains.…”

Exploring the Potential of Artificial Intelligence in Primary Care: Insights From Stakeholders’ Perspectives

“…XAI models for administrative tasks in areas other than healthcare have already begun to be explored such as: data management, record keeping, financial decision making, auditing, text classification and for hiring staff [67][68][69][70][71][72]. Indeed, further research in the areas of machine learning, Natural Language Processing (NLP), chatbots, and Robotic Process Automation (RPA) is warranted, particularly considering these AI approaches have been identified as valuable tools for handling administrative tasks within PC.…”

Innovating HRM Recruitment: A Comprehensive Review Of AI Deployment

“…The six significant theories that are deployed in algorithmic functioning are "ethical egoism-preference based, deontology-rule based, utilitarianism (principle-based), relativism, virtue ethics, and the ethics of care" (Rodgers et al, 2023). Hofeditz et al (2022) suggested that discrimination of AI may result from algorithms, but AI-based decision tools can make visible and retrievable errors; therefore, they can lead to more successful hires and reduced costs, eliminating unethical features and providing diversity in recruitment. The author deems that AI is unlikely to replace humans in recruiting procedures but rather to increase the role of humans to augment human decision-making for better outcomes.…”