Using AI-based detectors to control AI-assisted plagiarism in ESL writing: “The Terminator Versus the Machines”
Open accessThe DOI matches its Crossref record; 48 of the 49 references it cites were found in a registry or an index.
Abstract
The release of ChatGPT marked the beginning of a new era of AI-assisted plagiarism that disrupts traditional assessment practices in ESL composition. In the face of this challenge, educators are left with little guidance in controlling AI-assisted plagiarism, especially when conventional methods fail to detect AI-generated texts. One approach to managing AI-assisted plagiarism is using fine-tuned AI classifiers, such as RoBERTa, to identify machine-generated texts; however, the reliability of this approach is yet to be established. To address the challenge of AI-assisted plagiarism in ESL contexts, the present cross-disciplinary descriptive study examined the potential of two RoBERTa-based classifiers to control AI-assisted plagiarism on a dataset of 240 human-written and ChatGPT-generated essays. Data analysis revealed that both platforms could identify AI-generated texts, but their detection accuracy was inconsistent across the dataset.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures














Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“This bias is not incidental but structurally embedded in the design and training of these systems (Caner, 2025; Ibrahim, 2023; Jiang et al., 2024; Liang et al., 2023).”
From Harms and Ethical Risks of AI Detection Tools in Education: False Positives, Bias, Surveillance, and Student Rights (2026) · cites this paper as Ibrahim, 2023
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“These flaws weaken detection tools' credibility and call into question their admissibility as evidence in academic conduct proceedings [11].”
From Evaluating the Effectiveness and Ethical Implications of AI Detection Tools in Higher Education (2025) · cites this paper as [11]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“While AI detectors offer scalable solutions, they frequently produce false positives and lack transparency, especially for multilingual or non-native English speakers.”
From Evaluating the Effectiveness and Ethical Implications of AI Detection Tools in Higher Education (2025) · cites this paper as [11]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Non-native English students are most at risk for such mistakes because of the similarities in syntax and structure with AI-written language.”
From Evaluating the Effectiveness and Ethical Implications of AI Detection Tools in Higher Education (2025) · cites this paper as [11]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Inconsistencies were partly a result of the fact that both detectors were trained on GPT-2 output and were hence less adept at identifying content created by more advanced models like GPT-3.5 or GPT-4 [11].”
From Evaluating the Effectiveness and Ethical Implications of AI Detection Tools in Higher Education (2025) · cites this paper as [11]
Experimental — accuracy not yet measured
8 more claims about this paper
Create a free account to see every claim other papers make about this one.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805 · Cited by 73,173 · cited 1×
“An example of a zero-shot classifier is Gehrmann et al. (2019)’s Giant Language Model Test Room (GLTR), which utilizes BERT (Devlin et al., 2018) and GPT-2 (Radford et al., 2019) models to discriminate between machine-generated and human-written texts based on the probability of word distribution…”
MethodLanguage Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372 · cited 3×
“OpenAI used Common Crawl to train the model on a dataset of a trillion words of internet texts (175 billion parameters) in supervised training (Brown, 2020) and then fine-tuned the model using reinforcement learning from human feedback approaches (Christiano et al., 2017) to improve output…”
BackgroundMethodAn introduction to ROC analysis
Fawcett, T. (2006) · Pattern Recognition Letters · doi:10.1016/j.patrec.2005.10.010 · Cited by 21,879 · cited 1×
“After the scores for both datasets were compared for each detector to assess its ability to identify machine-generated texts, a confusion matrix (Fawcett, 2006) was developed to get a more accurate estimate of each detector’s accuracy.”
MethodRoBERTa: A Robustly Optimized BERT Pretraining Approach
Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., Levy, O., Lewis, M., Zettlemoyer, L., & Stoyanov, V. (2019) · arXiv · doi:10.48550/arxiv.1907.11692 · Cited by 16,515 · cited 4×
“In addition, the high effectiveness of RoBERTa-based detectors supports the conclusions of Liu et al. (2019) and Solaiman et al. (2019) that the optimized training of RoBERTa model results in significant performance improvements compared to other detection models.”
BackgroundMethodResultInfluentialAdvances in neural information processing systems 7
(1996) · Computers & Mathematics with Applications · doi:10.1016/s0898-1221(96)90463-0 · Cited by 14,281
Training language models to follow instructions with human feedback
Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C. L., Mishkin, P., Zhang, C., Agarwal, S., Slama, K., Ray, A., et al. (2022) · arXiv · doi:10.48550/arxiv.2203.02155 · Cited by 4,208 · cited 3×
“5 is a 175 billion parameter model (OpenAI, n.d.-a) trained using a combination of unsupervised machine learning and reinforcement learning from human feedback approaches to produce outputs aligned with users’ intents (Ouyang et al., 2022).”
BackgroundMethodChatting and cheating: Ensuring academic integrity in the era of ChatGPT
Cotton, D. R. E., Cotton, P. A., & Shipway, J. R. (2024) · Innovations in Education and Teaching International · doi:10.1080/14703297.2023.2190148 · Cited by 3,087 · cited 6×
“…and sudden rise of ChatGPT as a powerful generative technology marks a new age of AI-assisted plagiarism that constitutes an existential threat to higher education as it brings into question the reliability of its assessment practices and could devaluate the college degree (Cotton et al., 2023).”
BackgroundInfluentialChatting and Cheating. Ensuring academic integrity in the era of ChatGPT
Cotton, D., Cotton, P., & Shipway, J. R. (2023) · doi:10.35542/osf.io/mrz8h · Cited by 3,079
Chatting about ChatGPT: how may AI and GPT impact academia and libraries?
Lund, B. D., & Wang, T. (2023) · Library Hi Tech News · doi:10.1108/lhtn-01-2023-0009 · Cited by 1,355
Qualitative and descriptive research: Data type versus data analysis
Nassaji, H. (2015) · Language Teaching Research · doi:10.1177/1362168815572747 · Cited by 1,260 · cited 1×
“Generally speaking, descriptive research is well-suited for investigating L2 teaching and assessment since controlled experimental research is challenging to achieve in classroom settings (Nassaji, 2015).”
Background
Research works printed but not linked: 3
- 2. Baker, C. (2017). Quantitative research designs: Experimental, quasi-experimental, and descriptive. In H. Hall and L. Roussel (Eds.), Evidence-based practice: An integrative approach to research, administration, and practice, (2 nd Ed., pp.155-183).
- 18. Francke, E., & Bennett, A. (2019). The Potential Influence of Artificial Intelligence on Plagiarism: A Higher Education Per- spective. In P. Griffiths and M. N. Kabir (Eds.), European Conference on the Impact of Artificial Intelligence and Robotics (ECIAIR 2019) (pp. 131-140). Academic Conferences and Publishing Limited. DOI: https:// doi. org/ 10. 34190/ ECIAIR. 19. 043 · doi:10.34190/eciair.19.043
- 45. Radford, A., Wu, J., Child, R., Luan, D., Amodei, D., & Sutskever, I. (2019). Language models are unsupervised multitask learners. OpenAI Blog,1(8), 9.
Cited by
Items 101–104 of 104
“YOUR TEXT LOOKS LIKE FROM GPT”: AN ETHICAL STUDY OF AUTHORSHIP IN THE AGE OF LARGE LANGUAGE MODELS
Karnaukh, A. (2026) · Ezikov svât/Ezikov svât · doi:10.37708/ezs.swu.bg.v24i3.5
Liao, H., & Mohammed, L. A. (2026) · International journal of Emerging Issues in Social Science Arts and Humanities · doi:10.60072/ijeissah.2026.v4i04.004
Presumption of guilt? AI chatbots and the role of affirmation vivas in assessing academic integrity
Armstrong, L., & Karyotis, G. (2026) · AI and Ethics · doi:10.1007/s43681-026-01505-5
Khabirova, Z. (2026) · Research Square · doi:10.21203/rs.3.rs-11627257/v1
Metadata
- Authors
- Ibrahim, K.
- Affiliations
- Gulf University for Science & Technology (KW)
- Corresponding author
- Ibrahim, Karim
- ORCID
- Ibrahim, Karim
- Venue
- Language Testing in Asia
- Published
- 2023-10-16
- Volume
- 13
- Issue
- 1
- ISSN
- 2229-0443 (ISSN-L)
- OpenAlex
- W4387671615
- Cited by
- 66
- References
- 40
- Citations by year
- 2026: 14 · 2025: 28 · 2024: 22 · 2023: 2
- FWCI
- 16.55
- Type
- journal article
- Language
- en
- Update notices
- correction (2023-10-31)
- Open access
- gold
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0 (tdm)
- creativecommons.org/licenses/by/4.0 (vor)
- Locations
- Language Testing in Asia (publishedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- NA
- Topics
- Social Sciences › Education › Academic integrity and plagiarism
- Social Sciences › Education › Artificial Intelligence in Education
- Physical Sciences › Computer Science › Artificial Intelligence › Authorship Attribution and Profiling
- Keywords
- ChatGPT, ESL writing, writing assessment, AI-generated text detection, RoBERTa
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.
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.