When Humans and AI Begin to Think Alike: Cognitive Convergence and Decision Homogenization in AI-Augmented Management
Year: 2026
Type: preprint
Abstract: Artificial intelligence is increasingly embedded in managerial decision-making, where generative systems can formulate problems, synthesize evidence, generate alternatives, and recommend courses of action. Existing research has largely examined whether these systems improve individual decision quality, productivity, trust, or human-AI collaboration. Comparatively less attention has been given to how repeated exposure to AI-generated reasoning may influence the diversity and independence of human judgment itself. This conceptual article introduces human-AI cognitive convergence, defined as the progressive alignment of human problem framing, assumptions, reasoning patterns, interpretations, and solution preferences with those repeatedly generated or reinforced by artificial intelligence. It further proposes decision homogenization as a multilevel consequence through which individually converging judgments produce increasingly similar managerial interpretations and decisions across teams and organizations. The framework distinguishes cognitive convergence from agreement, automation bias, cognitive offloading, groupthink, and cognitive debt, and differentiates functional from dysfunctional convergence. It theorizes that AI-first interaction, repeated exposure, task ambiguity, perceived AI authority, and common-model use increase convergence, whereas domain expertise, independent-first reasoning, deliberate dissent, heterogeneous information sources, alternative framing, and multi-model challenge preserve cognitive diversity. The article develops an AI diversity paradox: systems capable of generating numerous alternatives may nevertheless reduce organizational cognitive diversity when many decision makers rely on similar AI-generated representations of what is relevant, reasonable, or actionable. The consequences of decision homogenization are expected to become most significant under novel, ambiguous, or weak-signal conditions requiring heterogeneous interpretations and unconventional responses. The article concludes with implications for human-AI governance and a research agenda for empirically examining cognitive convergence in AI-augmented managerial work. Keywords: artificial intelligence; generative AI; human-AI collaboration; cognitive convergence; decision homogenization; cognitive diversity; managerial decision-making; organizational adaptability; AI governance.
Author: Kutukoff, Sergey
Institution: Athabasca University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7225780411
Cites:
Cited by:
Related to: