From Cognitive Bias to Algorithmic Bias: A Neuro-Organizational Perspective on Recruitment Processes

Abstract

Cognitive bias in hiring decisions and algorithmic bias in artificial intelligence systems are largely treated as two independent research streams in the field of management and organization. This study establishes its primary theoretical objective as proposing an original conceptual framework that integrates these two streams and models the transmission mechanism between them. The central argument is based on the premise that algorithmic bias is not merely a technical problem, rather, human neurocognitive patterns are transferred to digital systems through individual decisions, institutional data structures, and algorithmic training processes, and are subsequently reproduced by these systems. Within this framework, the study proposes the concept of “neuro-bias,” which unifies processes treated on separate planes in the current literature within a cross-level transmission mechanism. The article systematically defines the boundaries and originality of the concept, puts forward five testable formal propositions, presents a multi-level framework regarding how the concept can be investigated at individual, institutional, and algorithmic levels, develops a comparative positioning against alternative theoretical approaches, and proposes a four-layered Hybrid Neuro-Organizational Decision Architecture (HNoDA) model that includes automation bias protection. Consequently, the study posits that algorithmic bias is not a technical system flaw, but a multi-level phenomenon fueled by the interactions among individual decision patterns, institutional data structures, and algorithmic learning processes. The proposed neurobias approach and the Hybrid Neuro- Organizational Decision Architecture (HNoDA) provide an integrated theoretical framework for developing fairer, more transparent, and sustainable decision systems in recruitment processes.