Document Type : Research Paper
Authors
1
Assistant Professor, Department of Foundations of Education, Shiraz University, Shiraz, Iran.
2
“Master’s student, Department of Foundations of Education, Faculty of Educational Sciences and Psychology, Shiraz University, Shiraz, Iran”
10.30487/rwab.2026.2086816.1680
Abstract
Abstract
The aim of this study was to provide a comprehensive and balanced analysis of the role of generative artificial intelligence (AI) in academic research and writing and to identify its opportunities, challenges, and ethical considerations based on the existing literature. This study adopted a systematic review approach and incorporated selected components of the PRISMA reporting framework. A structured search was conducted across reputable national and international databases covering publications from 2019 to 2024. Of the 52 initially identified records, duplicate records were removed, followed by title and abstract screening and full-text assessment. Ultimately, 26 eligible sources, including 25 English-language articles and one Persian-language article, were selected for analysis. Data were analyzed qualitatively through a four-step process consisting of the extraction of initial codes, their classification into 11 axial codes, the organization of these codes into three overarching themes, and a comparative analysis across studies. The findings revealed that generative AI offers significant opportunities in three main areas: enhancing research productivity, improving the quality of academic writing, and facilitating access to scholarly resources. At the same time, several challenges were identified, including threats to academic integrity and the risk of plagiarism, the potential weakening of critical thinking, excessive reliance on AI tools, and concerns regarding the accuracy and reliability of algorithmic outputs. Comparative analysis of the reviewed studies identified three dominant perspectives within the literature opportunity-oriented, threat-oriented, and balanced approaches as well as four major conceptual tensions.
Introduction
Academic writing is a form of scientific communication that demands precision, clarity, and adherence to established writing conventions. This skill represents a key component of university education, with practice being the most effective path to its development. However, the complexity and ambiguity inherent in academic writing have created numerous educational and structural challenges. Many students and researchers perceive writing as a difficult task, stemming from inadequate training and the perceived gap between technical and creative writing. Furthermore, the marginalized role of writing comprehension in research and education has deepened these challenges (Chan & Hu, 2023; Harmawan et al., 2023). Lack of mastery over grammar and excessive emphasis on the writing process, among both students and faculty, sometimes results in the production of weak scientific texts.
Generative artificial intelligence has created significant transformations in research and education, such that its integration into the writing of research articles, while increasing productivity, has provided a platform for analyzing the impacts of this technology on scientific communication and authorship. In this regard, applications of AI in academic research, intellectual property policies, and the benefits and challenges of its use in university libraries have been examined (Steiger, 2024; Kobrossy et al., 2025; Ateriya et al., 2025; John et al., 2023). The rapid advancement of generative AI, particularly large language models like ChatGPT, has made possible the transformation of research processes and academic book writing. This technology has the capacity to revolutionize areas such as ideation, structural compression, and analysis of large text datasets to identify patterns, as well as the ability to assist faculty and researchers in drafting and editing specialized scientific texts—a development that creates opportunities and simultaneously presents emerging ethical considerations in the academic ecosystem (Steiger, 2024; Mabirizi et al., 2025).
Traditional methods of scientific research and book writing faced serious structural limitations, including slow library searching, manual data analysis, and limited access to new findings, which prolonged the process of publishing theories and academic reference books. However, the advent of new tools has changed this balance, as Steiger (2024) and Mabirizi et al. (2025) demonstrated in their systematic explorations of generative AI's impact on quality, efficiency, ethics, and innovation in higher education research, showing that this technology improves academic writing, facilitates data analysis, and accelerates literature reviews. These transformative tools, while making library operations more efficient and contributing to content production processes, also possess the potential to create serious disruptions in the academic ecosystem—challenges manifesting as emerging plagiarism, reduced text originality, and the dissemination of inaccurate or biased scientific information, placing significant ethical considerations before writers and academic institutions.
Natural language processing tools have demonstrated high capability in simplifying traditional tasks of editors and human reviewers by transforming peer review and academic publication processes. Through automated text processing, format checking, and even initial plagiarism detection, these technologies can generate coherent and structured outputs that provide accurate foundations for drafting and refining comprehensive texts and academic books (Doskaliuk et al., 2025; Wang et al., 2025; Chan & Hu, 2023; Ateriya et al., 2025). Widespread access to these tools has brought positive transformations to higher education and academic writing, including personalized feedback and improved instructional materials, while simultaneously generating serious debates about intellectual property policies, reduced text originality, and emerging ethical challenges (Zhou et al., 2023; Hamoda et al., 2025; Kobrossy et al., 2025).
Review of the research literature reveals that although numerous studies have addressed various aspects of AI in education, none have provided an integrated framework capable of analyzing opportunities, challenges, and ethical considerations holistically within the process of compiling comprehensive academic works. Furthermore, there exists a conceptual gap in offering practical solutions for responsible use of AI in academic book writing; most studies are either entirely appreciative or critical, without presenting a balanced, solution-oriented perspective. The literature also lacks comparative analysis and identification of convergent patterns across studies. Accordingly, this article, through comprehensive review of recent literature, analyzes the role of generative AI in transforming research processes and academic book writing, seeking to delineate its opportunities, challenges, and ethical considerations in scientific communication and academic writing.
Research Methodology
This research employed a systematic review approach following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines to ensure methodological rigor and transparency in source selection. A comprehensive search was conducted across reputable domestic and international databases, including Scopus, Google Scholar, Noormags, Civilica, and Elmnet, for sources published between 2019 and 2024. Persian and English keywords were strategically combined, including "Generative AI," "Academic Writing," "ChatGPT," "Large Language Models," "Higher Education," "AI Ethics," and "Scientific Writing," to ensure comprehensive coverage of the literature.
From an initial identification of 52 documents, after removing duplicates (44 documents), screening titles and abstracts (removing 9 documents), and reviewing full texts (removing 9 additional documents due to lack of full-text access, focus on elementary education, or lack of direct relevance to academic research and writing), 26 sources (25 English articles and 1 Persian article) were ultimately selected for final analysis. Inclusion criteria were based on direct relevance to the topic, publication in peer-reviewed academic journals, and publication within the specified timeframe. Non-academic sources, non-research reports, and notes were excluded from the review process.
Data analysis was conducted qualitatively using thematic analysis methodology in four systematic stages: extracting initial codes from the selected sources; categorizing these into 11 core codes representing emerging patterns; organizing core codes into three main themes; and conducting comparative analysis across studies. To ensure the quality, objectivity, and trustworthiness of the thematic analysis, inter-coder reliability was established through the "inter-coder agreement" method, whereby an independent researcher coded and themed 10 articles (approximately 40% of total data) independently. The inter-coder agreement index reached 87%, exceeding standard thresholds and confirming the reliability of the analysis process. In cases of minor disagreement, extracted codes were reviewed in joint sessions and consolidated through final consensus.
Discussion
Opportunities of Generative AI in Academic Research and Writing
The comprehensive literature analysis reveals that generative AI provides significant opportunities across three primary domains: increasing research productivity, improving scientific writing quality, and facilitating access to research resources. These opportunities, extensively documented across multiple studies, represent transformative potentials for academic work while simultaneously necessitating careful consideration of accompanying challenges.
Enhancing Research Productivity and Efficiency
Evidence demonstrates that beyond merely increasing the speed and volume of text production, generative AI fundamentally transforms the nature of the research process. Researchers can redirect their cognitive efforts from repetitive tasks toward conceptual analysis, synthesis, and the design of innovative research questions (Mabirizi et al., 2025; Shahbazi et al., 2025; Hamoda et al., 2025; Doskaliuk et al., 2025; Wang et al., 2025; Cheng et al., 2025; Ghotbi, 2023; El Hani et al., 2024; Dinçer, 2024; Hosseinzadeh, 2025). AI tools based on large language models enable rapid data analysis, facilitate academic writing, and significantly increase researcher efficiency by processing large volumes of articles and extracting scientific patterns. This shift allows researchers to devote greater attention to higher-order thinking, critical analysis, and the generation of novel insights rather than being consumed by mechanical writing and data processing tasks.
Improving Scientific Writing Quality
Generative AI tools, through large language models, organize and simplify intensive writing tasks such as reviewing articles, laboratory reports, and experimental research, reducing cognitive load and enabling students to focus on critical analysis and conceptual development rather than merely technical aspects of writing (Harmawan et al., 2023; Zhou et al., 2023; Afifah, 2024; Chan & Hu, 2023; Steiger, 2024; El Hani et al., 2024; Dinçer, 2024). These tools enhance writing and research quality through personalized feedback, improved instructional materials, automated data analysis, repetitive task automation, advanced data analysis, and strengthened interdisciplinary collaboration. They render university library operations more efficient while simultaneously raising ethical concerns about devaluation of human writing skills.
Facilitating Access to Research Resources
Generative AI has significantly impacted scientific communication and academic writing, revealing both opportunities and challenges in its integration into research related to citation, authorship, and intellectual property (Kobrossy et al., 2025; Hamoda et al., 2025; Doskaliuk et al., 2025; Ghotbi, 2023; Carobene et al., 2024; El Hani et al., 2024; Dinçer, 2024). AI enables intelligent search and retrieval of sources, research design, macro-data analysis, and facilitates interdisciplinary collaboration. The rapid advancement of AI has transformed domains including academic publishing, peer review, and the process of writing scientific articles, with language model-based tools enabling simplification of tasks previously performed by human editors and reviewers (Doskaliuk et al., 2025; Wang et al., 2025; Carobene et al., 2024; El Hani et al., 2024; Dinçer, 2024).
Challenges and Limitations of Generative AI in Academic Writing
Despite its significant opportunities, generative AI presents substantial challenges that demand careful attention. Thematic analysis identifies four primary challenge categories: threats to scientific originality and increased plagiarism risk; weakening of critical thinking and researcher creativity; over-reliance on intelligent tools; and limitations in algorithmic accuracy, errors, and trustworthiness.
Threats to Scientific Originality and Plagiarism
Concerns regarding the weakening of student writing skills, plagiarism, production of false information, and fabricated citations have intensified with AI integration. Over-reliance on these tools threatens scientific integrity, research content accuracy, and the credibility of research outputs, particularly in sensitive fields such as medicine and engineering, potentially creating fundamental challenges to the usefulness and trustworthiness of academic research (Harmawan et al., 2023; Zhou et al., 2023; Chan & Hu, 2023; Cheng et al., 2025). The widespread integration of AI in research and scientific publishing has raised critical ethical considerations including intellectual property, authorship originality, algorithmic bias, data privacy, and excessive dependence on automation. Studies emphasize AI's dual role as both a tool for enhancing productivity and quality, and an ethical challenge requiring policy regulation, institutional training, and maintaining balance between technological capabilities and the cultivation of critical thinking and human creativity (Ghotbi, 2023; Lund et al., 2023; Carobene et al., 2024; Ersöz & Engin, 2024; Ugoala, 2025; Chekhratova & Pohorielova, 2024; Dinçer, 2024; Stahl et al., 2023).
Weakening of Critical Thinking and Creativity
Unlike challenges related to scientific originality primarily concerning research outcomes, the weakening of critical thinking pertains to learning processes and the formation of academic identity. Evidence indicates that generative AI leads some students and researchers to underestimate the necessity of developing writing and thinking skills, operating under the assumption that technology can replace their analytical and creative capabilities. This increasing dependence on AI tools, combined with algorithmic bias, data privacy risks, and excessive automation, can undermine critical thinking, human judgment, and scientific originality (Harmawan et al., 2023; Chan & Hu, 2023; Steiger, 2024). Furthermore, inequality in access to institutional support and appropriate training limits equitable utilization of these technologies, exacerbating existing disparities within the scientific community. Studies emphasize concerns regarding authorship preservation, the unique character of academic works, scientific integrity, and trustworthiness of AI-generated content, highlighting the necessity of serious attention to ethical considerations in adopting these tools in scientific writing (El Hani et al., 2024; Ugoala, 2025; Dinçer, 2024).
Over-Reliance on Intelligent Tools
While AI tools have provided significant opportunities through enhanced efficiency, accuracy, and facilitation of processes including writing, peer review, and plagiarism detection, excessive reliance on these tools raises substantial ethical and methodological concerns. These concerns include weakened writing originality, reduced role of human judgment, algorithmic bias, and data privacy risks (Afifah, 2024; Doskaliuk et al., 2025; Wang et al., 2025; Chan & Hu, 2023). Moreover, studies emphasize that increasing dependence on AI can exacerbate existing inequalities in resource access and institutional support, preventing some researchers from effectively utilizing these technologies. Concerns about preserving the originality of academic works, scientific integrity, and trustworthiness of AI-generated content highlight the necessity of ethical considerations in adopting these tools (Ateriya et al., 2025; El Hani et al., 2024; Ugoala, 2025; Dinçer, 2024).
Algorithmic Accuracy, Errors, and Trustworthiness
Despite increased AI efficiency in writing and analyzing scientific research, limitations regarding accuracy, content errors, and algorithmic trustworthiness pose serious challenges to academic integrity. Research emphasizes the necessity of strengthening researchers' critical thinking and maintaining their full responsibility for scientific content originality, accuracy, and relevance, particularly given risks of false information production, algorithmic bias, and privacy violations (Ghotbi, 2023; Wang et al., 2025; Cheng et al., 2025; Hamoda et al., 2025). Furthermore, the development and widespread use of large language models without clear ethical frameworks can facilitate misuse and undermine human judgment, necessitating balance between AI capabilities and cultivation of independent reasoning, research creativity, and responsible guidance of research and development, especially for early-career researchers (Ghotbi, 2023; Carobene et al., 2024; Lund et al., 2023; El Hani et al., 2024; Ugoala, 2025; Dinçer, 2024).
Ethical and Legal Considerations
Thematic analysis identifies four core ethical and legal considerations: intellectual property and authorship; transparency in AI use; scientific responsibility of authors and researchers; and algorithmic bias and scientific equity.
Intellectual Property and Authorship
Rapid advancement of generative AI, particularly large language models including ChatGPT and GPT-4, has created capacities for transformation in research while simultaneously raising intellectual property challenges including plagiarism, fabricated data, and researcher over-dependence, threatening research originality and intellectual independence (Mabirizi et al., 2025; Kobrossy et al., 2025; Carobene et al., 2024). The fundamental challenge concerns the blurring boundary between tool and author in text production, particularly concerning authorship attribution, scientific responsibility, and preservation of unique academic identity (Ateriya et al., 2025; Ugoala, 2025). Studies emphasize common ethical concerns including algorithmic opacity, bias, censorship, privacy violations, false information production, and fabricated data—concerns particularly significant in scientific writing where source reliability and data accuracy constitute core academic integrity elements (Bjelobaba et al., 2024; Dinçer, 2024; Chekhratova & Pohorielova, 2024). Proposed ethical frameworks emphasize principles including transparency, humanizing decision-making, inclusivity, human-machine collaboration, continuous evaluation, and ongoing learning, positioning AI not as replacement but as complement to human judgment and creativity (Eacersall et al., 2024; Stahl et al., 2023).
Transparency in AI Use
Given the dynamic and evolving nature of AI technologies, ongoing assessment of their ethical, scientific, and educational implications is essential. When used responsibly, generative AI can increase researcher productivity, provided that adherence to ethical standards and scientific integrity is maintained (Bjelobaba et al., 2024; Eacersall et al., 2024; Steiger, 2024). Studies emphasize the necessity of developing transparent institutional regulations and policies to address ethical and privacy concerns, as the absence of clear frameworks can enable unethical use by academic and research institutions (Chan & Hu, 2023; Ghotbi, 2023). While these technologies can increase efficiency, effectiveness, and quality of scientific articles, excessive reliance may weaken independent reasoning, creativity, and human judgment, particularly among early-career researchers (Carobene et al., 2024; Lund et al., 2023).
Scientific Responsibility and Authorship Accountability
Responsible use of generative AI requires the development of clear ethical frameworks and guidelines within academic institutions to ensure technological advancements are implemented inclusively, transparently, and accountably. These frameworks enable researchers to utilize AI capacities for increased efficiency and accuracy while maintaining scientific responsibility, research integrity, and human accountability (Eacersall et al., 2024; Wang et al., 2025; Hamoda et al., 2025). Qualitative and review research demonstrates that AI use in academic writing and publishing requires fundamental rethinking of ethical principles including intellectual property, scientific accuracy, transparency, and privacy protection. Challenges including algorithmic bias, data risks, excessive automation reliance, and diminished human judgment can undermine academic integrity, particularly where inequality in institutional support and training limits equal access (Ersöz & Engin, 2024; El Hani et al., 2024; Dinçer, 2024; Ugoala, 2025).
Algorithmic Bias and Scientific Equity
Concerns regarding algorithmic bias center on potential AI system discrimination based on training data characteristics, including gender, race, and socioeconomic biases inherent in training datasets, leading to distortion of scientific results and publication of biased academic texts (Ateriya et al., 2025; Ugoala, 2025; Eacersall et al., 2024). Inequality in access to advanced technologies and appropriate training creates and perpetuates digital divides, limiting equitable participation in AI-assisted research. Studies emphasize that while AI can enhance academic writing for those with institutional support and technological access, it may simultaneously widen gaps for researchers in under-resourced settings (Hamoda et al., 2025; Wang et al., 2025; Cheng et al., 2025; Chekhratova & Pohorielova, 2024). Addressing algorithmic bias requires development of diverse training datasets, transparent reporting of AI use, and institutional commitment to equitable access. Furthermore, researchers must critically evaluate AI-generated outputs to identify and correct biases.
Conclusion
This systematic review provides a comprehensive and balanced analysis of generative AI's role in academic research and writing, revealing a multi-dimensional phenomenon simultaneously offering significant opportunities and substantial challenges. The integration of generative AI into academic work represents not a binary choice between full acceptance or complete rejection, but rather demands a balanced, evidence-based, and critically informed approach. Key opportunities include enhanced research productivity, improved writing quality, and facilitated access to resources. However, these benefits are counterbalanced by serious challenges including threats to scientific originality, weakened critical thinking, over-reliance on AI tools, and algorithmic limitations regarding accuracy and trustworthiness.
Crucially, ethical and legal considerations—intellectual property and authorship, transparency, scientific responsibility, and algorithmic equity—emerge not as peripheral concerns but as central components in evaluating AI's implications for academic work. Comparative analysis of the literature reveals three prevailing approaches: opportunity-focused perspectives emphasizing practical benefits; threat-focused perspectives highlighting fundamental risks; and balanced approaches advocating responsible, conditional use of AI with transparency, human oversight, and ethical frameworks. Despite disagreements on specific aspects, virtually all studies converge on a fundamental principle: human responsibility, transparency in AI tool use, and strengthened ethical oversight throughout the research process are non-negotiable requirements.
Generative AI possesses significant capacity to transform academic research and writing, potentially increasing the speed, efficiency, and quality of knowledge production. However, effective utilization requires understanding limitations and ethical-legal challenges. AI's role must be defined as "assistive tool" rather than replacement for researcher judgment, creativity, and decision-making. Achieving this requires developing transparent institutional policies, education in ethical and digital literacy, strengthening critical thinking skills, and continuous human oversight. Through adherence to these frameworks, research opportunities can be maximized while preventing the weakening of scientific originality, trust in knowledge, and academic integrity. The study's findings, while robust, must be interpreted with consideration of several limitations, including restriction to Persian and English sources, variation in study methodologies, the rapid pace of AI technology evolution, and limited empirical studies in the Iranian higher education context. Future research should prioritize empirical investigations, particularly within specific disciplinary contexts, and examine the practical implementation of ethical frameworks in diverse academic settings.
Keywords
Subjects
