Paper Number
ICIS2026-2643
Paper Type
Short
Abstract
Efforts to institutionalize trustworthy Artificial Intelligence (AI) rely on governance mechanisms, such as human-in-the-loop, human-on-the-loop, human-in-command, which establish formal human oversight of automated decision-making (ADM). While these mechanisms provide foundational role-based oversight, oversight may be assigned to a designated human actor, with stakeholder perspectives entering, where considered, through consultation, representation, or surrounding organizational and regulatory arrangements. What remains less theorized is how plural stakeholder participation can function as an integral governance logic through which diverse perspectives and collective oversight are systematically incorporated into ADM governance. This study conceptualizes such logic within a generalizable AI governance mechanism. Drawing on stakeholder theory, it develops Stakeholders-in-the-loop (SITL), structuring plural oversight through asynchronous collection of evaluative, instructive, and authoritative stakeholder input, and its integration into ADM monitoring, supervision, and control. SITL complements existing role-based mechanisms by making stakeholder participation and collective governance an explicit oversight logic.
Recommended Citation
Göksal, Şaban İbrahim and Nikiforova, Anastasija, "From Human-in-the-Loop to Stakeholdersin-the-loop: A Multi-Stakeholder Governance Mechanism for AI Decision Making" (2026). ICIS 2026 Proceedings. 5.
https://aisel.aisnet.org/icis2026/policy/policy/5
From Human-in-the-Loop to Stakeholdersin-the-loop: A Multi-Stakeholder Governance Mechanism for AI Decision Making
Efforts to institutionalize trustworthy Artificial Intelligence (AI) rely on governance mechanisms, such as human-in-the-loop, human-on-the-loop, human-in-command, which establish formal human oversight of automated decision-making (ADM). While these mechanisms provide foundational role-based oversight, oversight may be assigned to a designated human actor, with stakeholder perspectives entering, where considered, through consultation, representation, or surrounding organizational and regulatory arrangements. What remains less theorized is how plural stakeholder participation can function as an integral governance logic through which diverse perspectives and collective oversight are systematically incorporated into ADM governance. This study conceptualizes such logic within a generalizable AI governance mechanism. Drawing on stakeholder theory, it develops Stakeholders-in-the-loop (SITL), structuring plural oversight through asynchronous collection of evaluative, instructive, and authoritative stakeholder input, and its integration into ADM monitoring, supervision, and control. SITL complements existing role-based mechanisms by making stakeholder participation and collective governance an explicit oversight logic.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
21_Policy_and_regulations