AI evidence extraction

Ethical Challenges of Generative AI in Academic Writing

Kartikkumar Pandya · International Journal on Science and Technology · 2025

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

9/10

Relevance

1/4

Quality (LMQS)

Evidence

Citations

0.00

FWCI

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.71097/ijsat.v16.i4.8913

Methodology & findings

Main result

The study found that "results of this study generally suggest a perceived increase in accessibility and efficiency in academic writing, thanks to personalization and help with difficult assignments. However, concerns were voiced about possible overreliance on the AI's output and academic dishonesty." Additionally, the research indicates "a dual nature to LLM usage: on one hand, they offer promising avenues for enhancing writing efficiency, supporting underserved students, and delivering timely feedback. Yet, on the other, they pose notable challenges related to plagiarism, over-reliance, and a potential decline in critical thinking abilities."

Reports effect sizes.

Research paradigm

Mixed-methods (qualitative and quantitative emphasis stated but not empirically executed)

Author conclusions

The authors conclude that "LLMs have been underscored as something of a two-prong approach, creating new ways of thinking about knowledge creation, while also complicating the ever-important question of academic integrity." They further assert that "there's a very real need for institutions to put in place policies and processes for framing how generative AI is used in education" and emphasize that "continued conversation is needed to develop equitable and responsible approaches as higher education institutions consider the ensuing changes that AI will be making to the educational landscape."

Risk of bias

No empirical data collection - all findings derive from literature review rather than primary research; No control conditions or comparison groups; No random sampling described; Potential selection bias in cited literature (may overrepresent available/published sources); No discussion of potential funding bias in reviewed studies; The paper does not explicitly identify specific bias risk factors in empirical research. However, it discusses the potential for AI models to "exhibit biases present in their training data, leading to biased outputs" and notes concerns about "reinforcement of existing societal biases."; No empirical data collection described; appears to be narrative literature review only; Selection bias in literature reviewed not specified; No mention of systematic search strategy or database selection criteria; Potential confirmation bias in interpretation of cited literature

Limitations

The paper acknowledges that "there remains a need for systematic empirical studies on the long-term effects of AI usage on student learning and faculty assessment practices" and notes "the limited research of the long-term implications of generative AI and the implications for academic integrity and students' learning experiences, with a prevailing concern academic integrity and the reliability and rigour of assessment systems." Additionally, the authors state that "most academic settings are unprepared for ethical governance."

Open questions raised

Lack of systematic empirical studies on long-term effects of AI usage on student learning and faculty assessment practices
Limited research on the long-term impacts of generative AI on student learning outcomes, such as writing quality and creativity
Need for studies addressing concerns about unequal access due to socioeconomic disparities
Scant research on the long-term implications of generative AI and implications for academic integrity and students' learning experiences
Need for future studies investigating the impact of LLMs on students' writing skills and academic integrity over time
The authors identify multiple research gaps: (1) need for systematic empirical studies on long-term effects of AI usage on student learning and faculty assessment practices; (2) limited research on long-term implications of generative AI for academic integrity and student learning experiences; (3) need for longitudinal studies examining impact of LLMs on writing skills and academic integrity over time; (4) research on frameworks for human-AI writing partnerships; (5) studies investigating opportunities and challenges of LLM accessibility across different student demographics; (6) need for interdisciplinary agreement on frameworks to mitigate ethical challenges of generative AI incorporation; (7) studies examining specific disciplinary applications of generative AI tools.

Extracted from: pdfAgreement 65%

Explore related topics

AI Manuscript Writing Academic Writing Qualitative