Publication Type

Journal Article

Version

submittedVersion

Publication Date

2-2025

Abstract

Generative AI (GenAI) has made rapid inroads in assessment, as a growing number of applicants rely on it as a coach in unproctored assessments of various selection procedures. This had led to assertions that applicants' GenAI use undermines key assumptions of the predictive model underlying selection and is thus disruptive for organizations' current unproctored assessments, thereby invoking various strategies of organizations to deter and detect its use. In this provocation article, we present a more nuanced view. To this end, we start by reviewing the recent research related to the effects of applicants' use of GenAI in assessment and discuss the evidence of the potential of applicant GenAI use to disrupt assessment validity. Next, we draw on test coaching frameworks to discuss three scenarios of how applicants' use of GenAI might affect an assessment's mean scores and criterion-related validity. These perspectives highlight that the use of GenAI might not only exert negative consequences but potentially have also positive consequences for both applicants and organizations. It is pivotal to distinguish among these scenarios because they lead to different strategies for organizations to deal with applicant use of GenAI.

Keywords

assessment, generative AI, large language models, selection, validity

Discipline

Human Resources Management

Research Areas

Organisational Behaviour and Human Resources

Publication

International Journal of Selection and Assessment

Volume

33

Issue

1

First Page

1

Last Page

9

ISSN

0965-075X

Identifier

10.1111/ijsa.12516

Publisher

Wiley

Copyright Owner and License

Authors

Additional URL

https://doi.org/10.1111/ijsa.12516

Download

Share

COinS