Paper Type

Complete

Abstract

Employees face opacity in how decisions are generated as they engage with AI systems. When such AI systems enter organizational decision-making, employees must work with outputs they cannot fully trace while remaining accountable for the outcomes. This creates a fundamental challenge before the AI system is even implemented. Drawing on a qualitative case study of 17 employees involved in HR decision processes in a large automotive corporation preparing to introduce an AI-based skill-matching system we adopt a relational affordance lens to investigate how perceived opacity shapes generative drive in the pre-implementation phase. Our findings show that perceived opacity constrains how employees anticipate and enact system affordances. It reduces epistemic confidence yet preserves personal accountability, destabilizing generative fit. Rather than suppressing, employees respond in two distinct ways: exploratory intensification or procedural safeguarding. These findings highlight the pre-implementation phase as formative and extend research on AI opacity and generative drive in organizations.

Paper Number

1828

Comments

SIG CNOW

Download

Share

COinS
 
Aug 15th, 12:00 AM

What Drives Us? How AI Opacity Shapes Generative Drive

Employees face opacity in how decisions are generated as they engage with AI systems. When such AI systems enter organizational decision-making, employees must work with outputs they cannot fully trace while remaining accountable for the outcomes. This creates a fundamental challenge before the AI system is even implemented. Drawing on a qualitative case study of 17 employees involved in HR decision processes in a large automotive corporation preparing to introduce an AI-based skill-matching system we adopt a relational affordance lens to investigate how perceived opacity shapes generative drive in the pre-implementation phase. Our findings show that perceived opacity constrains how employees anticipate and enact system affordances. It reduces epistemic confidence yet preserves personal accountability, destabilizing generative fit. Rather than suppressing, employees respond in two distinct ways: exploratory intensification or procedural safeguarding. These findings highlight the pre-implementation phase as formative and extend research on AI opacity and generative drive in organizations.