Detailed Information

Cited 0 time in webofscience Cited 0 time in scopus
Metadata Downloads

FAT-CAT—Explainability and augmentation for an AI system: A case study on AI recruitment-system adoption

Authors
Lee, ChangHyunCha, KyungJin
Issue Date
Mar-2023
Publisher
ACADEMIC PRESS LTD- ELSEVIER SCIENCE LTD
Keywords
Explainability; Human-AI augmentation; Technology adoption; AI system; Digital transformation
Citation
INTERNATIONAL JOURNAL OF HUMAN-COMPUTER STUDIES, v.171, pp.1 - 12
Indexed
SCIE
SSCI
SCOPUS
Journal Title
INTERNATIONAL JOURNAL OF HUMAN-COMPUTER STUDIES
Volume
171
Start Page
1
End Page
12
URI
https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/186075
DOI
10.1016/j.ijhcs.2022.102976
ISSN
1071-5819
Abstract
Because artificial intelligence (AI) recruitment systems exhibited discriminatory decisions in recent applications, the adoption of such systems in industry has raised doubts. As equity has been emphasized in AI decision-making frameworks, the non-explainability issue regarding the high performance of AI methods has become prominent. Therefore, scholars have focused on human–AI augmentation in which humans consider equity and AI supports the consideration. As a result, explainability is highlighted as a new capability of AI methods for an ideal decision. In this regard, this study proposes the so-called fairness, accountability, and transparency (FAT)-complexity, anxiety, and trust (CAT) model that describes the path from explainability to AI system adoption considering augmentation, assuming that the capability of the AI decision maker to explain the basis of its decision and interact with the human decision maker is crucial for AI recruitment system adoption. We found that explainability and augmentation are two key factors in AI recruitment system adoption and assessed that their importance will gradually increase as recruiters will be asked to use such AI systems more commonly. Moreover, this study conceptualized the role of an augmented relationship between humans and AI in decision-making, in which they complement each other's limitations.
Files in This Item
Go to Link
Appears in
Collections
서울 경영대학 > 서울 경영학부 > 1. Journal Articles

qrcode

Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.

Related Researcher

Researcher Cha, Kyungjin photo

Cha, Kyungjin
SCHOOL OF BUSINESS (SCHOOL OF BUSINESS ADMINISTRATION)
Read more

Altmetrics

Total Views & Downloads

BROWSE