AI evidence extraction

Using ChatGPT in academic writing is (not) a form of plagiarism: What does the literature say?

Adeeb M. Jarrah, Yousef Wardat, Patrícia Fidalgo · Online Journal of Communication and Media Technologies · 2023

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

207

Citations

7.38

FWCI

Top 10%

Impact

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.30935/ojcmt/13572

Methodology & findings

Study design

Systematic Literature Review (SLR) based on PRISMA guidelines.

Main result

The study found that "ChatGPT can be a valuable writing tool; however, it is crucial to follow responsible practices to uphold academic integrity and ensure ethical use. Properly citing and attributing ChatGPT's contribution is essential in recognizing its role, preventing plagiarism, and upholding the principles of scholarly writing." The review of 20 included studies revealed dual perspectives: some argue that using ChatGPT without proper citation constitutes plagiarism, while others contend that if generated content is critically evaluated, rephrased, and properly cited, its use is acceptable and does not constitute plagiarism.

Research paradigm

Interpretive/Hermeneutic

Author conclusions

The authors conclude that "while ChatGPT falls short of producing academic writing that meets the required academic journal publication standards, it excels in providing fast information with excellent language proficiency, primarily free of grammatical errors." They emphasize that "Maintaining academic integrity and avoiding plagiarism is paramount, and authors must be diligent in distinguishing between their original work and AI-generated text. Emphasizing the importance of responsible and ethical use of ChatGPT in academic writing is vital in academia." Furthermore, "By carefully considering the implications, implementing appropriate policies, and fostering mentorship and collaboration, institutions can empower students while safeguarding against academic dishonesty."

Risk of bias

Publication bias: only peer-reviewed journal articles included, excluding grey literature; Language bias: articles only in English language included; Selection bias in study inclusion based on specific keywords and database availability; Potential bias in included studies regarding ChatGPT's capabilities and limitations; Selection bias: Limited to peer-reviewed English-language articles only; Publication bias: Only published studies included, no grey literature; Database coverage bias: Specific databases selected may not capture all relevant literature; Language bias: Non-English publications excluded; Temporal bias: Search conducted fall 2022-spring 2023, may miss recent developments; Selection bias: Only peer-reviewed articles in English were included, potentially excluding relevant work in other languages or formats; Publication bias: Grey literature and unpublished studies were excluded; Timing bias: Study conducted during rapid emergence of ChatGPT; literature may not represent comprehensive evidence base; Screening bias: Four-stage selection process with author-determined criteria may introduce subjective filtering at title/abstract stage; Database coverage bias: Search limited to specific academic databases; may miss relevant publications in other sources; Author interpretation bias: Categorization of studies into application/benefit/risk categories was author-determined without inter-rater reliability reported

Limitations

The study acknowledges that "it is essential to understand that these technologies have limitations
They are not error-proof and should be used with human skill and judgment." Additionally, the authors note that "ethical considerations, transparency, and student privacy should be carefully addressed when implementing AI technologies in educational settings." The review is further limited by the fact that the literature on this emerging topic was nascent during the study period, and "the boundaries and guidelines for using these models are still being explored and defined within the academic community."

Open questions raised

Need for advanced detection methods to identify AI-generated content and plagiarism violations
Requirement for clear institutional policies on AI integration in educational settings
Development of guidelines for peer review processes to assess AI-augmented research
Addressing bias in AI models trained on datasets with biased information
Establishing consensus on authorship definitions and attribution when AI contributes to content creation
Bridging the digital divide to ensure equitable access to AI tools across regions and income levels

Extracted from: pdfAgreement 83%

Explore related topics

Citation Integrity Research Integrity Literature Review