Who Gets Seen in the Age of AI? Adoption Patterns of Large Language Models in Scholarly Writing and Citation Outcomes

2025-09-10

AI Summary

This study investigates how the global adoption of generative AI in academic writing reshapes scholarly visibility and citation dynamics, particularly across geopolitical regions. Method: Leveraging zero-shot AI detection, large-scale bibliometric analysis, and network science—complemented by a pre- versus post-ChatGPT comparative design—we examine adoption patterns, citation outcomes, and journal-level reception of AI-assisted manuscripts. Contribution/Results: We find higher AI adoption rates among scholars in Eastern regions, yet Western scholars accrue disproportionately larger citation gains due to preexisting academic capital. Top-tier journals exhibit implicit stylistic preferences for “human-like” prose, resulting in lower acceptance rates for AI-assisted texts—a systemic “use–reward mismatch.” Critically, this is the first study to demonstrate that generative AI tools are not epistemically neutral: their diffusion exacerbates, rather than alleviates, regional scholarly inequities, revealing how structural power asymmetries mediate technological adoption in academia.

Technology Category

• Philosophy and Ethics of AI: Bias, Fairness & Equity
• Humans and AI: AI for Accessibility
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy

Application Category

• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• Web Mining and Content Analysis: Web data quality in the era of algorithmically-generated content

Abstract

The rapid adoption of generative AI tools is reshaping how scholars produce and communicate knowledge, raising questions about who benefits and who is left behind. We analyze over 230,000 Scopus-indexed computer science articles between 2021 and 2025 to examine how AI-assisted writing alters scholarly visibility across regions. Using zero-shot detection of AI-likeness, we track stylistic changes in writing and link them to citation counts, journal placement, and global citation flows before and after ChatGPT. Our findings reveal uneven outcomes: authors in the Global East adopt AI tools more aggressively, yet Western authors gain more per unit of adoption due to pre-existing penalties for "humanlike" writing. Prestigious journals continue to privilege more human-sounding texts, creating tensions between visibility and gatekeeping. Network analyses show modest increases in Eastern visibility and tighter intra-regional clustering, but little structural integration overall. These results highlight how AI adoption reconfigures the labor of academic writing and reshapes opportunities for recognition.

Problem

• Examining AI adoption disparities in scholarly writing across regions
• Analyzing citation outcomes linked to AI-assisted writing styles
• Investigating visibility and gatekeeping tensions in academic publishing

Innovation

• Zero-shot detection of AI-likeness
• Network analyses of citation flows
• Stylistic changes tracking methodology