What question did this study set out to answer?

This paper aims to challenge the assumption that AI can fully replace human labor in organizations.

June 20, 2026Open Access

The Automation Efficiency Paradox: Why Generative AI Increases the Need for Human Expertise

Key Points

This paper aims to challenge the assumption that AI can fully replace human labor in organizations.
Introduced a behavioral and cognitive framework to analyze the effects of using AI in organizations.
Evaluated the implications of conversational AI outputs on human decision-making and cognitive processes.
Examined the relationship between AI use and the need for human expertise in operational workflows.
AI outputs often lead to cognitive fatigue in human operators, resulting in flawed data acceptance.
The system's design can camouflage errors instead of eliminating them, leading to potential mistrust.
AI increases the demand for skilled human experts to verify and guide algorithmic outputs, preserving workflow integrity.

Abstract

Organizations are rapidly adopting Large Language Models (LLMs) under the assumption that automation can seamlessly replace human labor to maximize operational speed (Srinivasan, 2026). This paper introduces the Automation Efficiency Paradox to challenge this assumption using a behavioral and cognitive framework (Bashkirova Perez et al., 2022). Rather than eliminating human error, the system camouflages omissions behind masterfully crafted, fluid prose (Srinivasan, 2026). When scaling tasks to large, macro-level text blocks, human operators experience cognitive fatigue and cave to premature verification convergence, unthinkingly accepting flawed data (Elsevier B.V., 2026). We argue that blindly replacing human workers with AI leads to a downstream collapse of trust in the system's output (Bashkirova Elsevier B.V., 2026).

Read Full Paperexternally

Ask AI

Helpful

Bookmark

View Full Paper