AI evidence extraction

Impact of Large Language Models on Research Writing and Publication Practices

Devinder Kaur · International Journal of Sciences and Innovation Engineering · 2026

AI-generated evidence extraction, verified across multiple analytical personas. Not a substitute for the peer-reviewed original.

9/10

Relevance

Evidence

Citations

0.00

FWCI

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.70849/ijsci03022699663

Methodology & findings

Study design

Narrative literature review synthesizing existing research and institutional policies on LLM use in academic writing; no primary data collection or empirical testing.

Main result

Large Language Models are transforming academic research writing by improving linguistic clarity, productivity, and accessibility for researchers. The paper notes that "LLMs can generate coherent text, revising drafts, improving grammar, and summarizing complex documents" and that "when used responsibly, they can assist scholars in expressing ideas more effectively without compromising the intellectual content of their work." However, integration also raises important concerns about academic integrity, authorship responsibility, and the risk of hallucinated content.

Research paradigm

interpretivist/constructivist

Author conclusions

"Large Language Models (LLMs) are transforming academic research writing by improving linguistic clarity, productivity, and accessibility for researchers. When used responsibly, they can assist scholars in expressing ideas more effectively without compromising the intellectual content of their work." The authors conclude that "LLMs should be viewed as assistive rather than substitutive tools, with clear disclosure, ethical adherence, and sustained human oversight being essential to ensuring they strengthen, rather than undermine, the credibility of academic scholarship."

Risk of bias

Publication bias - review may over-represent published studies favorable to LLM use; Selection bias - non-systematic search strategy not explicitly described; Reporting bias - reliance on secondary sources from publishers and editorial organizations; No original empirical data presented; review relies on citations to existing literature; Potential publication bias in cited sources (no mention of searching grey literature or unpublished studies); Narrative structure may reflect author selection bias in which studies and perspectives are included; Limited empirical data on actual impact quantification; Policy-based rather than evidence-based conclusions

Open questions raised

Future research directions include: improved factual accuracy in LLMs, reduced hallucinated content, stronger alignment with academic norms, ongoing research in model training and fine-tuning, development of standardized policies for LLM use in academia, and educational initiatives to help researchers understand appropriate roles of LLMs in academic writing.

Extracted from: pdf

Explore related topics

AI Manuscript Writing Academic Writing Literature Review