An Analytical Scoping Review of How Artificial Intelligence Reliance Deprives Academics of Creative Synthesis

Abstract

Background: Generative artificial intelligence (AI) is now embedded in academic reading, writing, and ideation. Its benefits for speed and fluency are well publicised. Its costs to the higher-order intellectual act that defines scholarship, creative synthesis, are far less examined. Creative synthesis means combining distant ideas, tolerating uncertainty, and producing a new conceptual arrangement that neither the sources nor the scholar held beforehand.

Objective: This paper analyses how habitual reliance on AI systems can deprive academics of the conditions under which creative synthesis develops, and it appraises the strength of the evidence for this claim.

Methods: A PRISMA-informed scoping review was conducted in two strands. Strand A was a transparent, title-level screening of the 117-record corpus supplied for this study. Strand B was a purposive search for theoretical and empirical anchor studies in cognitive psychology, creativity research, and AI ethics. Records were charted against five mechanisms of deprivation.

Results: Twenty-nine corpus records and eighteen external sources were retained (n = 47). Evidence converged on five mechanisms: cognitive offloading, the loss of productive difficulty, automation complacency, output homogenisation, and epistemic narrowing. Experimental evidence shows that AI assistance raises individual output quality while reducing the diversity of ideas across a group, and that performance gains during assisted work do not always survive its removal.

Conclusions: The evidence does not show that AI use is inherently harmful. It shows that unstructured, substitutive reliance erodes the effortful integrative processes on which synthesis depends. The paper closes with an evidence-graded framework for scholars and institutions.