Cited by 424Artificial Intelligence in EducationArtificial Intelligence in Healthcare and EducationAcademic integrity and plagiarism

Analyzing the role of ChatGPT as a writing assistant at higher education level: A systematic review of the literature

Open access
Imran, Muhammad, Almusharraf, Norah·2023·Contemporary Educational Technology

Abstract

This study examines the role of ChatGPT as a writing assistant in academia through a systematic literature review of the 30 most relevant articles. Since its release in November 2022, ChatGPT has become the most debated topic among scholars and is also being used by many users from different fields. Many articles, reviews, blogs, and opinion essays have been published in which the potential role of ChatGPT as a writing assistant is discussed. For this systematic review, 550 articles published six months after ChatGPT’s release (December 2022 to May 2023) were collected based on specific keywords, and the final 30 most relevant articles were finalized through PRISMA flowchart. The analyzed literature identifies different opinions and scenarios associated with using ChatGPT as a writing assistant and how to interact with it. Findings show that artificial intelligence (AI) in education is a part of the ongoing development process, and its latest chatbot, ChatGPT is a part of it. Therefore, the education process, particularly academic writing, has both opportunities and challenges in adopting ChatGPT as a writing assistant. The need is to understand its role as an aid and facilitator for both the learners and instructors, as chatbots are relatively beneficial devices to facilitate, create ease and support the academic process. However, academia should revisit and update students’ and teachers’ training, policies, and assessment ways in writing courses for academic integrity and originality, like plagiarism issues, AI-generated assignments, online/home-based exams, and auto-correction challenges.

AI snapshot

Written from the paper’s full text.

TL;DR

This systematic review analyzed 30 articles published December 2022–May 2023 to assess ChatGPT as a writing assistant in higher education. It finds ChatGPT offers opportunities as a facilitator for drafting, brainstorming, and feedback, but raises challenges for accuracy, academic integrity, authorship, and assessment.

Key findings

  • From 550 initially identified articles, 30 were included after PRISMA screening; more than 70% of the main statements from selected literature highlighted ChatGPT’s role as a writing tool/assistant.
  • ChatGPT can assist with generating text, initial drafts, brainstorming ideas, literature summaries, grammar/syntax feedback, and writing structure, but it cannot replace human writers.
  • Concerns include unreliable or erroneous outputs, fabricated references/DOIs, lack of latest information (ChatGPT-4 limited to information before September 2021), plagiarism/originality issues, and disapproval of ChatGPT as a co-author.
  • Most selected literature was published in medical journals; the USA produced more articles than other countries, with the UK and Australia also contributing.

Methods

  • PRISMA-guided systematic review searching Scopus, Science Direct, PubMed, and Web of Science using keyword and Boolean search strings across identification, screening, eligibility, and inclusion stages.
  • From 550 articles, 25 duplicates and 321 non-criteria articles were excluded, leaving 204 for eligibility; 174 more were excluded, yielding 30 final articles.
  • Analysis summarized background information (countries, journals, citations, subject areas) and main statements/capabilities from the selected documents in tables; no meta-analysis was performed.

Limitations

  • Stated: the review only covers literature on ChatGPT as a writing assistant at higher education level, not other functions; quality and depth of reviewed literature may vary; rapidly evolving NLP means the literature may not capture the latest developments; higher-education focus may limit transferability to other contexts.
  • Evident: the synthesis rests on only 30 articles from a six-month window (December 2022–May 2023), mostly from medical journals and English-language indexed sources, which may narrow the range of perspectives.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.30935/cedtech/13605

Cites

53 of 54 references verified0 citation errors · 1 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 55 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    This study found that text generated by ChatGPT and proposed searched material, including references and other publication features like authors' names and DOIs, contain errors.

    “However, this study found that text generated by this AI chatbot and proposed searched material, including references and other publication feathers like authors' names and DOIs, contain errors.”
    Quote in the paperSubstantial claim
  • Findingreasoning

    ChatGPT can facilitate collaboration among students and educators by allowing multiple people to work on a project simultaneously.

    “ChatGPT can also facilitate collaboration among students and educators.Using AI to generate content allows multiple people to work on a project simultaneously, allowing for more efficient collaboration.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Geher, 2023

      Not checked in the free run.

    • cites Holmes et at., 2022

      Not checked in the free run.

    • cites McMurtrie, 2022

      Not checked in the free run.

  • Findingcitation

    ChatGPT can help students generate new ideas for writing assignments by suggesting topics, themes, and perspectives they might not have considered otherwise.

    “ChatGPT can help students generate new ideas for their writing assignments by suggesting topics, themes, and perspectives that they might not have considered otherwise (Kasneci et al., 2023;Taecharungroj, 2023).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Kasneci et al., 2023

      Not checked in the free run.

    • cites Taecharungroj, 2023

      Not checked in the free run.

  • Findingown data

    More than 70% of the main statements from selected literature for this systematic review highlighted ChatGPT's use as a writing tool and writing assistant.

    “More than 70% (Table 4) of the main statements from selected literature for this systematic review highlighted this aspect.”
    Quote in the paperSubstantial claim
  • Interpretationcitation

    ChatGPT could work or assist as a powerful tool in academic and scientific writing.

    “In this systematic review, the role of ChatGPT as a writing assistant has been analyzed as an AI tool based on the generative pre-trained transformer (GPT) that could work/assist as a powerful tool in academic and scientific writing.”
    Quote in the paperSubstantial claim
  • May combine two claims. 42
  • Classified as probably not a substantial claim. 57
  • May combine two claims.

    Findingcitation

    Even ChatGPT accepted those errors, expressed regret for them, and tried to produce similar results.

    “Even ChatGPT accepted those errors, expressed regret for them, and tried to produce similar results (Roumeliotis & Tselikas, 2023).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Roumeliotis & Tselikas, 2023

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    The findings gleaned from the literature reviewed and analysis suggest that ChatGPT can proffer, identify and collect data for research purposes to facilitate prospective researchers' work in getting relevant resources.

    “The findings gleaned from the literature reviewed and analysis suggest that ChatGPT can proffer, identify and collect data for research purposes to facilitate prospective researchers' work in getting relevant resources.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Findingcitation

    ChatGPT can proofread and edit student writing by suggesting corrections for grammar, syntax, and spelling errors, helping students improve writing quality and reduce errors.

    “ChatGPT can proofread and edit student writing by suggesting corrections for grammar, syntax, and spelling errors.This can help students improve the quality of their writing and reduce errors (Geher, 2023;Holmes et at., 2022;McMurtrie, 2022).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Geher, 2023

      Not checked in the free run.

    • cites Holmes et at., 2022

      Not checked in the free run.

    • cites McMurtrie, 2022

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Findingnone given

    This systematic review highlights a significant research gap in the utilization and ability of AI chatbots and other technologies, including ChatGPT, to tackle machine learning tasks in an efficient way.

    “Moreover, this systematic review highlights a significant research gap in the utilization and ability of AI chatbots and other technologies, including ChatGPT, to tackle machine learning tasks in an efficient way.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingcitation

    Zheng and Zhan (2023) observed the role of ChatGPT in academic and scientific writing and highlighted serious concerns.

    “The same issue Zheng and Zhan (2023) discussed when they observed the role of ChatGPT in academic and scientific writing and highlighted serious concerns.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Zheng and Zhan (2023)

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    Most articles appeared in medical journals, and only a few relevant and high-quality articles were published in arts and humanities and social sciences-related journals.

    “But most articles appeared in medical journals, and only a few relevant and high-quality articles were published in arts and humanities and social sciences-related journals.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Findingnone given

    There are many ethical issues with AI-generated writing, such as plagiarism and authentic data utilization, and the accuracy of writing products is still questionable.

    “There are many ethical issues, such as plagiarism and authentic data utilization, and the accuracy of writing products is still questionable.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

    Findingnone given

    AI in education is part of the ongoing development process, and its latest chatbot, ChatGPT, is part of it.

    “Findings show that artificial intelligence (AI) in education is a part of the ongoing development process, and its latest chatbot, ChatGPT is a part of it.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Few scholars recognized ChatGPT as a powerful writing assistant and added it as a co-author in their research articles.

    “Few other scholars (Kung et al., 2023;Manohar & Prasad, 2023;Transformer & Zhavoronkov, 2022) have also recognized ChatGPT as a powerful writing assistant and added it as a co-author in their research articles.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Kung et al., 2023

      Not checked in the free run.

    • cites Manohar & Prasad, 2023

      Not checked in the free run.

    • cites Transformer & Zhavoronkov, 2022

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Findingnone given

    No regulations, rules, or guidelines are currently available about the use of ChatGPT and other AI tools and issues related to plagiarism, copyright, authorship, and attribution for text generated by them.

    “No regulations, rules, or guidelines are currently available about its use and issues related to plagiarism, copyright, authorship, and attribution for the text generated by ChatGPT and other AI tools.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingcitation

    Zimmerman (2023) concluded that the best way to familiarize and get used to AI platforms is to use ChatGPT like other educational software and gadgets regularly, but AI tools still have certain limitations, especially for medical practitioners to get deeply involved in AI technologies for responsible uses in medical affairs.

    “Zimmerman (2023) evaluated the prospective role of ChatGPT as a genuine writer or a ghostwriter through a detailed analysis of AI and natural language processing (NLP) systems and concluded that the best way to familiarize and get used to these platforms is to use this chatbot like other educational software and gadgets regularly.However, this study concluded that AI tools still have certain”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Zimmerman (2023)

      Not checked in the free run.

  • May combine two claims.

    Findingreasoning

    ChatGPT can offer significant assistance in generating text, initial drafts, brainstorming ideas, and summaries of the literature, but it cannot replace or be considered like a human writer in various skills and knowledge.

    “The major findings of this study conclude that ChatGPT can offer significant assistance in generating text, initial drafts, brainstorming ideas, and summaries of the literature, but it cannot replace or be considered like a human writer in various skills and knowledge.”
    Quote in the paperSubstantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Many scholars disapprove of listing ChatGPT as a co-author and writing assistant in academic works, particularly research items.

    “However, according to Stokel-Walker (2023) and Thorp (2023), this idea is disapproved by many scholars to list ChatGPT as a co-author and writing assistant in academic works, particularly research items.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Stokel-Walker (2023)

      Not checked in the free run.

    • cites Thorp (2023)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingcitation

    Suaverdez and Suaverdez (2023) tested a few chatbots like ChatGPT, Google Bard, and generative AI for mapping their human-like responses in academic writing; despite their benefits, these chatbots have challenges of lack of originality, inaccuracies, too generalized responses, and poor logic flow in writing prompts.

    “They also tested a few chatbots like ChatGPT, Google Bard, and the generative AI for mapping their human-like responses in academic writing; however, despite their benefits, these chatbots have challenges of lack of originality, inaccuracies, too generalized responses, and poor logic flow in writing prompts.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Suaverdez and Suaverdez (2023)

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    ChatGPT is used as a writing tool/assistant in different subject areas, and authors have highlighted its role in creative, academic, and scientific writing as a facilitator and author.

    “ChatGPT is used as a writing tool/assistant in different subject areas, and authors have highlighted its role in creative, academic, and scientific writing as a facilitator and author.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Findingcitation

    ChatGPT has access to vast information, making it easier to identify and eliminate errors, leading to a higher likelihood of producing accurate and consistent content.

    “4. More accurate and consistent content: With the ChatGPT invention, there is a higher likelihood of producing accurate and consistent content.ChatGPT has access to vast information, making identifying and eliminating errors easier (Stacey, 2022).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Stacey, 2022

      Not checked in the free run.

  • May combine two claims.

    Findingcitation

    da Silva (2023) analyzed AI-generated passages and found them very causal and suspected to be plagiarized from actual sources with minor paraphrasing; therefore, ChatGPT is not considered much successful in providing authentic and pertinent literature.

    “Besides, da Silva (2023) also analyzed AI-generated passages and found them very causal and suspected to be plagiarized from actual sources with minor paraphrasing.Therefore, ChatGPT is not considered much successful in providing any authentic and pertinent literature.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites da Silva (2023)

      Not checked in the free run.

  • May combine two claims.

    Findingcitation

    ChatGPT can translate text from one language to another, which can help students writing in a non-native language ensure their writing is accurate and grammatically correct.

    “ChatGPT can translate text from one language to another, which can be useful for students who are writing papers in a language that is not their native tongue.This can help students ensure that their writing is accurate and grammatically correct (Lametti, 2022;Lund & Wang, 2023;Stock, 2023).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Lametti, 2022

      Not checked in the free run.

    • cites Lund & Wang, 2023

      Not checked in the free run.

    • cites Stock, 2023

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    Scholars in the USA produced more articles than other countries on ChatGPT and writing issues; UK and Australia were also among other countries that contributed.

    “The scholars in the USA produced more articles than other countries in disseminating the debate on ChatGPT and writing issues.UK and Australia were also among other countries that contributed to studies on ChatGPT as a writing assistant and an AI tool-related theme.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingcitation

    Altmae et al. (2023) shared their personal experience using ChatGPT as a writing assistant for scientific writing and found it a very helpful, convenient, and user-friendly tool.

    “Moreover, Altmae et al. (2023) shared their personal experience using ChatGPT as a writing assistant for scientific writing and found it a very helpful, convenient, and user-friendly tool.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Altmae et al. (2023)

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Findingown data

    The analyzed literature identifies different opinions and scenarios associated with using ChatGPT as a writing assistant and how to interact with it.

    “The analyzed literature identifies different opinions and scenarios associated with using ChatGPT as a writing assistant and how to interact with it.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The statements and findings reported in the reviewed articles are accurate and reliable.
  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingcitation

    After a huge debate across academia and publishing houses, almost all publishers and journals introduced policies about the status of ChatGPT as a co-author or facilitator.

    “However, after that, a huge debate across academia and publishing houses started, where publishers, editors, and researchers questioned the adoption of AI as a co-author instead of taking it as any other writing tool or AI technology.After this debate, almost all publishers and journals introduced their policies about the status of ChatGPT as a co-author or facilitator (Stokel-Walker, 2022).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Stokel-Walker, 2022

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    In the study by Yan (2023), participants generally showed more concern rather than satisfaction towards the unrestricted application of ChatGPT in L2 writing.

    “In the study, participants generally showed more concern rather than satisfaction towards the unrestricted application of ChatGPT in L2 writing”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Yan, 2023, p. 18

      Not checked in the free run.

  • May combine two claims.

    Findingcitation

    ChatGPT can reduce the time and effort required to generate written content, allowing students and educators to save time and focus on other aspects of their work.

    “1. Increased efficiency: ChatGPT's invention can reduce the time and effort required to generate written content.With its ability to generate coherent and well-structured text on any topic, students and educators can save time and focus on other aspects of their work (Lund et al., 2023;Yan, 2023).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Lund et al., 2023

      Not checked in the free run.

    • cites Yan, 2023

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    Among the thirty selected studies, most supported ChatGPT as an advanced AI generative model that can produce good responses to users' queries; however, no one can 100% rely on its responses due to various technical errors and the unavailability of the latest information.

    “Among thirty selected studies, most of them have supported ChatGPT as an advanced AI generative model that can produce good responses to users' queries; however, no one can 100% rely on its responses due to various technical errors and the unavailability of the latest information”
    Quote in the paperSubstantial claim
    • Rests on an assumption: ChatGPT-4 can only produce information from a limited time frame, like before September 2021.
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Authors of various publications from different countries have observed and highlighted that ChatGPT and its role as a writing tool/assistant received attention worldwide.

    “However, authors of various publications from different countries have observed and highlighted that ChatGPT and its role as a writing tool/assistant received attention worldwide (Figure 3).”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Findingcitation

    Studies by Dergaa et al. (2023), Dwivedi et al. (2023), and Lund (2023) pointed out significant differences between ChatGPT and human-generated writings, such as ChatGPT has the ability to produce concise abstracts, introductory passages, and literature sections, but produced limited responses compared to human content.

    “In (Dergaa et al. 2023;Dwivedi et al., 2023;Lund, 2023), the authors carried out studies to verify the use of ChatGPT in scientific writings to judge its content's quality, reliability, and comparative analysis with human-generated content.These studies further investigated AI-generated content and pointed out significant differences between ChatGPT and human-generated writings, such as ChatGPT”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Dergaa et al. 2023

      Not checked in the free run.

    • cites Dwivedi et al., 2023

      Not checked in the free run.

    • cites Lund, 2023

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationown data

    No one can mindlessly rely on the results produced and suggestions made by ChatGPT.

    “No one can mindlessly rely on the results produced and suggestions made.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationnone given

    ChatGPT's role in the fields of writing, research, and education can increase productivity and enhance efficiency.

    “Notably, ChatGPT's role in the fields of writing, research, and education can increase productivity and enhance efficiency.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

    Interpretationreasoning

    AI-generated writing will never replace human writing.

    “Undoubtedly, AI technology and tools have greatly advanced in recent times and revolutionized how academic and scientific writings work; however, there is a dire need to understand that AI-generated writing will never replace human writing.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Intelligent tutoring systems have a moderately positive impact on college students' academic achievement.

    “In a meta-analysis-based study of the effectiveness of ITS on college learners' academic performance, Steenbergen-Hu and Cooper (2014) concluded that ITS has a moderately positive impact on college students' academic achievement.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Steenbergen-Hu and Cooper (2014)

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductnone given

    For this systematic review, 550 articles published from December 2022 to May 2023 were collected based on specific keywords, and the final 30 most relevant articles were finalized through a PRISMA flowchart.

    “For this systematic review, 550 articles published six months after ChatGPT's release (December 2022 to May 2023) were collected based on specific keywords, and the final 30 most relevant articles were finalized through PRISMA flowchart.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 30 selected articles are representative of the published literature on ChatGPT as a writing assistant.
  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    Finally, 30 most relevant articles were considered eligible for further examination and analysis.

    “Finally, 30 most relevant articles were considered eligible for further examination and analysis.In the final round of inclusion, 30 documents were included that fulfilled the criterion of this systematic review.”
    Quote in the paperSubstantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationnone given

    Like other language correction tools such as Grammarly and Quill Bot, ChatGPT can also work as a writing support and feedback tool that will help to identify advanced grammar and syntax issues, suggestions for writing structure improvement, and coherence of the prompts.

    “Moreover, like other language correction tools such as Grammarly and Quill Bot, ChatGPT can also work as a writing support and feedback tool that will help to identify advanced grammar and syntax issues, suggestions for writing structure improvement, and coherence of the prompts.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    550 articles were identified and retrieved from the four major databases.

    “550 articles were identified and retrieved from the four major databases.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Interpretationcitation

    The development of AI technologies like ChatGPT can impact future writing tasks potentially.

    “The finding from the literature reviewed systematically highlights that the development of AI technologies like ChatGPT can impact future writing tasks potentially.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Study design and conductnone given

    The rapidly evolving nature of NLP technologies, like ChatGPT, means that the literature available at the time of the study may not fully capture the advancement and most recent developments in this field.

    “Moreover, the rapidly evolving nature of NLP technologies, like ChatGPT, means that the literature available at the time of the study may not fully capture the advancement and most recent developments in this field.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Over the past two decades, AIED research mostly focused on learner-supporting AI technologies that aim to make learners independent and automate teacher functions.

    “Over the past two decades, artificial intelligence and education (AIED) research mostly focused on learnersupporting AI technologies that aim to make learners independent and automate teacher functions (Afzaal et al., 2022).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Afzaal et al., 2022

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The primary focus of this study on higher education settings might limit the transferability of the findings to other educational contexts.

    “Finally, the primary focus of this study on higher education settings might limit the transferability of the findings to other educational contexts.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    This systematic review is conducted through PRISMA as a guide.

    “This systematic review is conducted through preferred reporting items for systematic review and meta-analysis (PRISMA) as a guide (Liberati et al., 2014).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Liberati et al., 2014

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    ChatGPT received more than 100 million active users just two months after its launch in November 2022.

    “Since its launch in November 2022, ChatGPT received huge success in January 2023, and it became one of the fastest and most highly welcomed artificial intelligence (AI) technological tools by open-AI, with more than 100 million active users just in two months (Williams, 2023).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Williams, 2023

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    Leading databases Scopus, Science Direct, PubMed, and Web of Science were systematically used to review literature in four stages: identification, screening, eligibility, and inclusion.

    “For this process, leading databases, Scopus, Science Direct, PubMed, and Web of Science (WoS), were systematically used to review literature in four stages: identification, screening, eligibility, and inclusion.”
    Quote in the paperSubstantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    25 duplicate publications were excluded from the 550 initially collected documents.

    “Screening process was done to check and remove duplicate publications from the selected databases, and 25 duplicate publications were excluded from 550 initially collected documents.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Altmae et al. (2023) are of the view that ChatGPT is an advanced AI tool, and researchers should use it very carefully by considering its pros and cons.

    “However, Altmae et al. ( 2023) are of the view that ChatGPT is an advanced AI tool, and researchers should use it very carefully by considering its pros and cons.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Altmae et al. (2023)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

    Study design and conductnone given

    The quality and depth of the literature reviewed may vary, potentially affecting the robustness and generalizability of the findings of this study.

    “Secondly, the quality and depth of the literature reviewed may vary, potentially affecting the robustness and generalizability of the findings of this study.”
    Quote in the paperNot a substantial claim
  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductcitation

    PRISMA is followed as a guide because it helps examine a large database of research articles, clearly addresses research questions, and enables identification of inclusion and exclusion criteria.

    “Throughout this systematic review, PRISMA is followed as a guide because it is commonly used in educational research projects due to its three advantages.First, it helps examine a large database of research articles and other literature.Second, it clearly addresses research questions for a smooth, systematic research process, and finally, it makes the researcher able to identify the inclusion and”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Ramalingam et al., 2022

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    This study only covers the existing literature on ChatGPT's role as a writing assistant at the higher education level, and all other functions and uses are not discussed in this study.

    “First, this study only covers the existing literature on ChatGPT's role as a writing assistant at the higher education level, and all other functions and uses are not discussed in this study.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    According to Firat (2023), ChatGPT and other AI technologies significantly impact students' academic performance and can potentially revolutionize traditional learning practices, foster soft skills development and promote personalized learning experiences.

    “According to Firat (2023), ChatGPT, including other AI technologies, significantly impact students' academic performance and can potentially revolutionize traditional learning practices, foster soft skills development and promote personalized learning experiences.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Firat (2023)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    The connection between AI and education is based on the solution to many core problems the education sector faces, such as lack of qualified teachers, virtual assistance, and the growing underachievement gap among rich, poor and mediocre learners.

    “The connection between AI and education is based on the solution to many core problems the education sector is facing, such as the lack of qualified teachers, virtual assistance, and the growing and underachievement gap among rich, poor and mediocre learners (Holmes et al., 2022;Imran & Almusharraf, 2023).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Holmes et al., 2022

      Not checked in the free run.

    • cites Imran & Almusharraf, 2023

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationnone given

    Development in AI technologies, including ChatGPT, can potentially revolutionize and challenge the existing writing techniques and domains and can have both challenges and opportunities, ranging from positive to negative aspects.

    “Ultimately, development in AI technologies, including ChatGPT, can potentially revolutionize and challenge the existing writing techniques and domains and can have both challenges and opportunities, ranging from positive to negative aspects.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    After screening, 204 documents were selected for the eligibility test.

    “After the screening, only 204 documents were selected for the eligibility test so that it could be determined through a rigorous examination of the titles, abstracts, keywords, and major contents to double-check the inclusion criteria, as mentioned in Table 2.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Interpretationreasoning

    The education process, particularly academic writing, has both opportunities and challenges in adopting ChatGPT as a writing assistant.

    “Therefore, the education process, particularly academic writing, has both opportunities and challenges in adopting ChatGPT as a writing assistant.”
    Quote in the paperSubstantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Alattar and McDowell (2023) believed that in the near future, most writing tasks would be accomplished by AI tools.

    “To advance the use of AI tools in scientific writing, Alattar and McDowell (2023) navigated the use of ChatGPT as a rising AI coauthor in various projects such as they believed that in the near future, most of the writing tasks would be accomplished by these AI tools.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Alattar and McDowell (2023)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

    Interpretationreasoning

    ChatGPT's best use would be strengthened as an assistant for researchers, teachers, and students to use it as a collaborative tool for receiving feedback, customizing search results, seeking suggestions, and identifying ethical and academic integrity considerations in the future.

    “ChatGPT's best use would be strengthened as an assistant for researchers, teachers, and students to use it as a collaborative tool for receiving feedback, customizing search results, seeking suggestions, and identifying ethical and academic integrity considerations in the future.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    321 documents that did not meet the criteria were excluded, such as publications before November 2022, non-English publications, and articles that did not deal with the determined keywords.

    “Moreover, 321 documents that did not meet the criteria were excluded, such as publications before November 2022, non-English publications, and articles that did not deal with the determined keywords.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    ChatGPT is an open-access and public tool that works according to GPT language model technology.

    “ChatGPT, being an open-access and public tool, is a highly sophisticated chatbot that works according to GPT language model technology (Kirmani, 2022, p. 574).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Kirmani, 2022, p. 574

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    This process strictly followed the PRISMA guidelines to ensure the selection of high-quality and reliable data for this study.

    “This process strictly followed the PRISMA guidelines to ensure the selection of high-quality and reliable data for this study.”
    Quote in the paperSubstantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    During the eligibility stage, 174 more documents were excluded because their abstracts and main contents did not meet the research objectives and were not related to ChatGPT and AI as writing assistant context.

    “So, during this eligibility stage, 174 more documents were excluded because their abstracts and main contents did not meet the research objectives and were not related to ChatGPT and AI as writing assistant context.”
    Quote in the paperNot a substantial claim
  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductown data

    Through the PRISMA flow chart, thirty articles were analyzed based on the primary search keywords: ChatGPT and AI in education, ChatGPT as a writing assistant, ChatGPT and the future of writing, and AI and scientific writing.

    “Through PRISMA flow chart, thirty articles were analyzed based on the primary search keywords: ChatGPT and AI in education, ChatGPT as a writing assistant, ChatGPT and the future of writing, and AI and scientific writing.”
    Quote in the paperSubstantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    ChatGPT can generate statistical analyses, lyrics, computer programs, and abstracts or introductions to scientific articles.

    “It can also generate statistical analyses, lyrics, computer programs, and abstracts or introductions to scientific articles”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Rahimi & Abadi, 2023

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation

    ChatGPT can generate high-quality, plausible, human-like written responses.

    “ChatGPT can generate high-quality, plausible, human-like written responses.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Rahimi & Abadi, 2023

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Conversational AI is likely to revolutionize research practices and publishing, creating both opportunities and concerns.

    “Conversational AI is likely to revolutionize research practices and publishing, creating both opportunities and concerns”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Van Dis et al., 2023

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Premisecitation

    ChatGPT is among the most transformative AI tools developed in recent years.

    “ChatGPT undoubtedly is among the most transformative AI tools developed in recent years”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Dwivedi et al., 2023

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Current natural language processing AI represent a significant threat to the fidelity of short-form essays as an assessment method in physics courses.

    “… current natural language processing AI represent a significant threat to the fidelity of short-form essays as an assessment method in physics courses”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Yeadon et al., 2023

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    ChatGPT is likely to be a useful tool for educators designing science units, rubrics, and quizzes.

    “ChatGPT is likely to be a useful tool for educators designing science units, rubrics, and quizzes”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Cooper, 2023

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    ChatGPT is a versatile and valuable tool with significant potential to promote engaging and adaptive language learning.

    “ChatGPT is a versatile and valuable tool with significant potential to promote engaging and adaptive language learning”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Kohnke et al., 2023

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Due to the phenomenon of hallucinations, ChatGPT users must carefully proofread output to ensure that they are accurate and ready for use.

    “Due to the phenomenon of hallucinations, ChatGPT users must carefully proofread output to ensure that they are accurate and ready for use”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Ge & Lai, 2023

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

    Premisecitation

    The potential positive and negative impacts of ChatGPT are immediate and wide-ranging.

    “The potential positive and negative impacts of ChatGPT are immediate and wide-ranging, emphasizing the necessity for a comprehensive examination and collective formation of ethical guidelines for its use in all fields”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Taecharungroj, 2023

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

    Premisecitation

    AI is already used in medicine, especially in image analysis, but the domains are infinite, and it is possible that AI could quickly help or replace rheumatologists in the writing of scientific articles.

    “AI is already used in medicine, especially in image analysis, but the domains are infinite, and it is possible that AI could quickly help or replace rheumatologists in the writing of scientific articles”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Verhoeven et al., 2023

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    ChatGPT can significantly enhance both the efficiency and the quality of writing review articles for scientists.

    “ChatGPT can significantly enhance both the efficiency and the quality of writing review articles for scientists”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Huang & Tan, 2023

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    ChatGPT is an AI software potentially able to assist in the writing process of a scientific paper and can help in the literature review, identify research questions, provide an overview of the current state of the field, and assist with tasks such as formatting and language review.

    “ChatGPT is an AI software potentially able to assist in the writing process of a scientific paper and can help in the literature review, identify research questions, provide an overview of the current state of the field, and assist with tasks, such as formatting and language review”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Salvagno et al., 2023

      Not checked in the free run.

71 of 71
  • The paper relies on this without support.

    The 30 selected articles are representative of the published literature on ChatGPT as a writing assistant.

    “Finally, 30 most relevant articles were considered eligible for further examination and analysis.”
  • The paper relies on this without support.

    The statements and findings reported in the reviewed articles are accurate and reliable.

    “Through PRISMA flow chart, thirty articles were analyzed based on the primary search keywords: ChatGPT and AI in education, ChatGPT as a writing assistant, ChatGPT and the future of writing, and AI and scientific writing.”
  • The paper relies on this without support.

    ChatGPT-4 can only produce information from a limited time frame, like before September 2021.

    “Like humans, ChatGPT cannot produce the latest information as current ChatGPT-4 can only produce information from a limited time frame, like before September 2021.”
  • The paper relies on this without support.

    Chatbots are relatively beneficial devices to facilitate, create ease and support the academic process.

    “as chatbots are relatively beneficial devices to facilitate, create ease and support the academic process”
4 of 4

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Other studies suggested that generative AI systems such as ChatGPT are increasingly being adopted to assist researchers with literature synthesis, information organization, and preliminary drafting.

    “Likewise, other studies suggested that generative AI systems such as ChatGPT are increasingly being adopted to assist researchers with literature synthesis, information organization, and preliminary drafting, reflecting a growing transition toward AI-assisted scholarly workflows [12, 40].”

    From Developing the HAKI model as a conceptual framework for AI-assisted academic research (2026) · cites this paper as [40]

    The “ChatGPT and Future of Writing” section says ChatGPT can help researchers with literature summaries, initial drafts, customized search results, and suggestions. It does not establish that researchers are increasingly adopting it, and it does not specifically describe information organization as a use. (LLM, conf=0.84)

    Experimental — accuracy not yet measured
  • ChatGPT can offer advice regarding various structural aspects of a text and translate it.

    “offer advice regarding various structural aspects of a text and translate it (Imran & Almusharraf, 2023)”

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Conflicting claims about ChatGPT authorship policies

    Major

    The paper says almost all publishers and journals introduced policies on ChatGPT's status as a co-author or facilitator, but later says no rules or guidelines are available on its use and authorship.

1 of 1

Cites

Flagged first, then by citations

Research works printed but not linked: 2

  1. Lund, B. D. (2023). A brief review of ChatGPT: Its value and the underlying GPT technology. https://www.researchgate.net/publication/366809571_A_Brief_Review_of_ChatGPT_Its_Value_and_the_Underlying_GPT_Technology
  2. Suaverdez, J. B., & Suaverdez, U. V. (2023). Chatbots impact on academic writing. Global Journal of Business and Integral Security.

Cited by

Sorted by citations

Authors
Imran, M., & Almusharraf, N.
Affiliations
Prince Sultan University (SA)
Corresponding author
Imran, Muhammad
Venue
Contemporary Educational Technology
Published
2023-10-01
Volume
15
Issue
4
Pages
ep464
ISSN
1309-517X (ISSN-L)
OpenAlex
W4386072877
Cited by
421
References
50
Citations by year
2026: 113 · 2025: 161 · 2024: 139 · 2023: 8
FWCI
17.39
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
Contemporary Educational Technology (publishedVersion, cc-by)
Funders
Prince Sultan University
Topics
  • Physical Sciences › Computer Science › Information Systems › Artificial Intelligence in Education
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
  • Social Sciences › Safety Research › Academic integrity and plagiarism
Keywords
ChatGPT, writing assistants, academic writing, artificial intelligence in education, chatbots, higher education, academic integrity, plagiarism, online examinations, automatic correction, writing instruction

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-06.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-06.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review