When the Curriculum Vitae meets the Algorithm

Lui, A ORCID: 0000-0003-2463-5177 (2025) When the Curriculum Vitae meets the Algorithm. In: Okur, S, (ed.) Artificial Intelligence Law and Digitalisation in Law. Science and Thought Series No: 65 . TUBA – Turkish Academy of Sciences, pp. 1-24. ISBN 9786256110663

Preview

Text When the Curriculum Vitae meets the Algorithm.pdf - Published Version Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (358kB) | Preview

Abstract

UK employers increasingly use algorithmic hiring systems. These systems are often more complex than algorithmic credit-scoring systems. The hiring process involves a wider range of data types (structured and deeply unstructured, including video and text) and requires sophisticated multimodal tools. Further, the process must account for nuanced ethical, social, and legal variables, and often blends human and artificial intelligence (Aİ) decision-making in a multi-layered workflow. This is complicated by the fact that some algorithmic scores are so influential that the boundary between scoring and automated decision-making is hard to establish. İt is well known that the ‘black box’ opacity phenomenon can reinforce and perpetuate algorithmic challenges such as bias, discrimination, and unfairness. The “black box” phenomenon in algorithmic hiring systems exacerbates biases even when protected characteristics such as age and race are excluded from the input data. Data which appear ‘neutral’, such as educational background or postcodes, may lead to bias and discriminatory outcomes. Certain postcodes may be associated with large populations of immigrants or ethnic minorities (European Union Agency for Human Rights, 2022). As a result, algorithms may unintentionally produce biased outcomes, particularly if the data used for training are not representative. This can cause harmful or unfair effects, such as bias or discrimination against certain job applicants, especially those from vulnerable groups, when making automated hiring decisions. This chapter critically reviews the challenges associated with automated hiring systems, especially at the screening and interview stages. İt critically analyses recent U.S. court decisions on automated hiring systems and associated problems such as algorithmic bias and discrimination. The issue of third-party liability of Aİ software vendors will also be evaluated, in light of U.S. and European Union (EU) cases and legislation. The chapter argues that employers must provide a transparency notice when using automated hiring systems to job applicants at the outset, increasing transparency and trust between humans and automated hiring systems. The concept of an attractive virtual agent or avatar is also recognised as a means to increase affective trust in Aİ technology. Finally, third-party Aİ software vendors should be mindful of their potential liability when automated Aİ decisions play a significant role in the recruitment process.

Item Type:	Book Section
Uncontrolled Keywords:	48 Law and Legal Studies; 4807 Public Law; Minority Health; Data Science; Machine Learning and Artificial Intelligence; Bioengineering; Networking and Information Technology R&D (NITRD); Social Determinants of Health; Health Disparities and Racial or Ethnic Minority Health Research; Health Disparities; 16 Peace, Justice and Strong Institutions
Subjects:	K Law > K Law (General) T Technology > T Technology (General)
Divisions:	Law and Justice Studies
Publisher:	TUBA – Turkish Academy of Sciences
Date of acceptance:	1 December 2025
Date of first compliant Open Access:	28 October 2026
Date Deposited:	28 Oct 2026 09:51
Last Modified:	28 Oct 2026 09:51
DOI or ID number:	10.53478/tuba.978-625-6110-66-3
Editors:	Okur, S
URI:	https://researchonline.ljmu.ac.uk/id/eprint/29753

View Item