The DOI matches its Crossref record; 17 of the 17 references it cites were found in a registry or an index.
Do teachers spot AI? Evaluating the detectability of AI-generated texts among student essays
Open accessAbstract
The potential application of generative artificial intelligence (AI) in schools and universities poses great challenges, especially for the assessment of students’ texts. Previous research has shown that people generally have difficulty distinguishing AI-generated from human-written texts; however, the ability of teachers to identify an AI-generated text among student essays has not yet been investigated. Here we show in two experimental studies that novice (N = 89) and experienced teachers (N = 200) could not identify texts generated by ChatGPT among student-written texts. However, there are some indications that more experienced teachers made more differentiated and more accurate judgments. Furthermore, both groups were overconfident in their judgments. Effects of real and assumed source on quality assessment were heterogeneous. Our findings demonstrate that with relatively little prompting, current AI can generate texts that are not detectable for teachers, which poses a challenge to schools and universities in grading student essays. Our study provides empirical evidence for the current debate regarding exam strategies in schools and universities in light of the latest technological developments.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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?
In the citing paper
“Previous studies suggested that human users often fail to recognise AI-generated text and are generally overconfident in their ability to do so (Casal & Kessler, 2023;Fleckenstein et al., 2024).”
From “ ChatGPT can make mistakes. Check important info.” Epistemic beliefs and metacognitive accuracy in students' integration of ChatGPT content into academic writing (2025) · cites this paper as Fleckenstein et al., 2024
In the citing paper
“This is heightened by the fact that even experts, such as professional linguists and teachers, are unable to effectively distinguish between human-written and AI-generated text (Casal & Kessler, 2023;Fleckenstein et al., 2024).”
From “ ChatGPT can make mistakes. Check important info.” Epistemic beliefs and metacognitive accuracy in students' integration of ChatGPT content into academic writing (2025) · cites this paper as Fleckenstein et al., 2024
Citations
Cites
ChatGPT for Good? On Opportunities and Challenges of Large Language Models for Education
Kasneci, E., Seßler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · doi:10.35542/osf.io/5er8f · Cited by 7,552
ChatGPT for good? On opportunities and challenges of large language models for education
Kasneci, E., Sessler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · Learning and Individual Differences · doi:10.1016/j.lindif.2023.102274 · Cited by 7,326
Chatting 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
Chatting 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
Gao, C. A., Howard, F. M., Markov, N. S., Dyer, E. C., Ramesh, S., Luo, Y., & Pearson, A. T. (2022) · doi:10.1101/2022.12.23.521610 · Cited by 1,099
Köbis, N., & Mossink, L. D. (2021) · Computers in Human Behavior · doi:10.1016/j.chb.2020.106553 · Cited by 406
Readers’ perception of computer-generated news: Credibility, expertise, and readability
Graefe, A., Haim, M., Haarmann, B., & Brosius, H.-B. (2018) · Journalism · doi:10.1177/1464884916641269 · Cited by 315
Meyer, J., Jansen, T., Schiller, R., Liebenow, L. W., Steinbach, M., Horbach, A., & Fleckenstein, J. (2024) · Computers and Education: Artificial Intelligence · doi:10.1016/j.caeai.2023.100199 · Cited by 311
Self-citationSailer, M., Bauer, E., Hofmann, R., Kiesewetter, J., Glas, J., Gurevych, I., & Fischer, F. (2023) · Learning and Instruction · doi:10.1016/j.learninstruc.2022.101620 · Cited by 106
Keller, S., Fleckenstein, J., Krüger, M., Köller, O., & Rupp, A. A. (2020) · Journal of Second Language Writing · doi:10.1016/j.jslw.2019.100700 · Cited by 65
Self-citation
Cited by
Sorted by citations
Items 101–159 of 159
Silang, C. D. (2026) · QualiSearch Journal of Educational Research and Practice · doi:10.66437/qjerp.v2si.017
Musical Giftedness in Human-AI Creation: Interpreting Evidence Beyond the Finished Product
García-Díaz, I. (2026) · Gifted Education International · doi:10.1177/02614294261498145
Kaya, B., & Yaman, S. A. (2026) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2026.2747530
Fajri, J., Rofi'udin, A., & Harsiati, T. (2026) · Paedagogia : jurnal pendidikan · doi:10.24239/pdg.vol15.iss2.1477
Zaitsu, W., Jin, M., Ishihara, S., Tsuge, S., & Inaba, M. (2026) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2026.1918939
Liu, S., & Qin, Y. (2026) · Acta Psychologica · doi:10.1016/j.actpsy.2026.107940
Yastıbaş, A. E., Dincer, A., & Darıyemez, T. (2026) · Online Learning In Educational Research (OLER) · doi:10.58524/zphezp55
“LLMs’ ability to generate convincing natural-language responses (Stokel-Walker, 2023 ) seems to have improved GenAI’s ability to mimic human writing, making it harder for teachers to identify who or what produced the text (Cotton et al., 2024; Fleckenstein et al., 2024).”
Jauhiainen, J. S. (2026) · Academia AI and Applications · doi:10.20935/acadai8542
Sadowski, B. M., & Oliveira, M. (2026) · Assessment in Education Principles Policy and Practice · doi:10.1080/0969594x.2026.2736625
AI-enabled Influence Operations in the Indo-Pacific: The Philippine Perspective
Gomez, M. A., & Mount, G. (2026) · Journal of Asian Security and International Affairs · doi:10.1177/23477970261485092
Detecting the undetectable? Reassessing academic misconduct procedures in the era of generative AI
Teichmann, F. M. J. (2026) · Assessment & Evaluation in Higher Education · doi:10.1080/02602938.2026.2728036
Munoz, A., Hinchcliff, M., Langfield, C. T., & Rogerson, A. M. (2026) · International Journal for Educational Integrity · doi:10.1007/s40979-026-00235-9
“Controlled experiments demonstrate that both novice and experienced teachers are unable to identify AI-generated text among student submissions, and that both groups are overconfident in their judgements (Fleckenstein et al. 2024).”
Ethical Use of Artificial Intelligence in Research Publication: A Scoping Review
Dakhesh, S., Khademizadeh, S., Atlasi, R., Zareivenovel, M., & Ouchi, A. (2026) · International Journal of Health Policy and Management · doi:10.34172/ijhpm.9759
“21 The use of AI-generated content in academic work may threaten scholarly integrity and has been linked to increased student plagiarism.”
Susoy, Z. (2026) · Pamukkale University Journal of Education · doi:10.9779/pauefd.1847553
Kosztyó, P., Burkovics, M., & Jenei, P. (2026) · Journal of Physics Conference Series · doi:10.1088/1742-6596/3280/1/012028
“Only 45.1% óf pré-sérvicé téachérs córréctly distinguishéd AI-génératéd éssays fróm human-writtén ónés in óné study [12], and 51.21% óf judgés cóuld córréctly différéntiaté bétwéén AI-génératéd and human-writtén póétry in anóthér [13].”
InfluentialWhen the panic sets in: AI-proofing as pedagogical violence in social work education
Hosken, N., & Bailey, S. (2026) · Social Work Education · doi:10.1080/02615479.2026.2691573
Söğüt, S., & BÜYÜKKIDIK, S. (2026) · Educational Assessment Evaluation and Accountability · doi:10.1007/s11092-026-09495-y
Contemplative Literacy in the Age of GenAI: Embodied Reading, Empathy, and the Limits of Detection
Smith, C. (2026) · CEA critic · doi:10.1353/cea.2026.a994419
Villacura-Herrera, C., Zaal, F. T. J. M., & Cox, R. F. A. (2026) · Computers in Human Behavior Reports · doi:10.1016/j.chbr.2026.101221
Detecting and Visualizing LLM Usage in Students' Essays Leveraging Past Assignments and LLM Answers
Zhao, Y., Lueder, R., Zhang, Q., & Starner, T. E. (2026) · ACM Conference on Learning @ Scale (L@S) · doi:10.1145/3774398.3811559
“However, this approach performs less reliably on edited or partially LLM-assisted content [1].”
Rethinking the role of teachers in generative AI-supported task-based language teaching
Sun, Q.-H. (2026) · Language Teaching · doi:10.1017/s0261444826101372
Detección de plagio y autoría en la era de los LLM: precisión y efectos pedagógicos
Salazar, L. M., & Guaña–Moya, J. (2026) · Ethos Scientific Journal · doi:10.63380/esj.v4n1.2026.352
Humans vs. machines: can enaction theory help?
Pallanti, L., & Chaker, R. (2026) · AI-Linguistica Linguistic Studies on AI-Generated Texts and Discourses · doi:10.62408/ai-ling.v4i1.50
“…number of teachers and trainers, both in secondary education and in higher education, struggle to ascertain whether the texts they evaluate constitute original student productions or outputs generated by generative artificial intelligence (GenAI) (Fleckenstein et al. 2024; Nguyen and Barrot 2024).”
Gultom, I., Puspadini, R., Erwis, F., Prasiwiningrum, E., & Ridwan (2026) · JOURNAL OF ICT APLICATIONS AND SYSTEM · doi:10.56313/jictas.v5i1.531
Tashmeer, M., Davies, M. J., & Stephens, J. M. (2026) · British Educational Research Journal · doi:10.1002/berj.70208
Villalobos, D. J. M. (2026) · Estudios y Perspectivas Revista Científica y Académica · doi:10.61384/r.c.a..v6i2.1911
Text features associated with students’ generative AI use: Norwegian teachers’ perspectives
Jeffery, J. V., Skar, G. B. U., Kvistad, A. H., & Graham, S. (2026) · Frontiers in Education · doi:10.3389/feduc.2026.1792351
"It's OK Because...": The Wild West of Student Rationalization of AI Use in Academic Writing
Kim, J., Toyama, K., Kim, S., & Carroll, J. M. (2026) · arXiv.org · doi:10.48550/arxiv.2605.29090
“This causes significant challenges for instructors to assess students’ essays (Fleckenstein et al. 2024), while students are uneasy to not use AI for various reasons like flexibility of AI, social norms, or time efficiency (Adnin et al. 2025; Gi-ray et al. 2026) Extensive research has examined how…”
Posibilidades y retos del uso del ChatGPT en la escritura argumentativa en el nivel universitario
Mosquera, E. M., Sierra, L. O., & Hernández-Henríquez, M.-M. (2026) · Revista Virtual Universidad Católica del Norte · doi:10.35575/rvucn.n78a8
AI and the Reduction of Truth: A Eucharistic Alternative
Reilly, C. M. (2026) · Religions · doi:10.3390/rel17050551
Muthmainnah, N., Risdianto, F., Muslim, A., Sholichah, A. L. A., & Ayu, E. W. N. (2026) · FOSTER Journal of English Language Teaching · doi:10.24256/foster-jelt.v7i2.324
“The integrity of students in using AI tools for academic writing is often called into question (Bouteraa et al., 2024; Fleckenstein et al., 2024; Roe & Perkins, 2022).”
M, I. M. I., Fandini, I., Ampry, E. S., Hasbahuddin, H., Syarah, E., Resa, A., Nova, S. T., Prawansyah, A., Nurpadila, N., Sulistiyawati, S., et al. (2026) · GERVASI Jurnal Pengabdian kepada Masyarakat · doi:10.31571/gervasi.v10i1.9543
“(Fleckenstein et al., 2024) juga menambahkan bahwa pemanfaatan AI dapat mengurangi hambatan dalam proses menulis sehingga produktivitas meningkat.”
Regional Language Interference in Indonesian Phonology among Elementary School Students
Nahak, K. B., Nahak, M. M. N., Lasi, F., Seran, Y., Lein, A. L., & Verdial, J. D. C. (2026) · Jurnal Penelitian dan Pengembangan Pendidikan · doi:10.23887/jppp.v10i1.102894
Educomunicación para la interpretación de información financiera en la toma de decisiones
Ocádiz-Amador, V., Ocádiz-Amador, A., & González-Ruiz, O. (2026) · COMMERCIUM PLUS · doi:10.53897/revcommerp.2026.08.04
Granja, M. E. A., Pirca, C. E. C., Lagla, E. G. C., Culqui, J. I. C., & Achig, M. F. H. (2026) · Ibero-American Journal of Education & Society Research · doi:10.56183/iberoeds.v6i1.717
“An experiment that was developed by (Fleckenstein et al., 2024) consisted of testing whether teachers with and without experience were able to identify a text generated by AI and a text made by a student, the study reveals that the two groups of teachers presented difficulties that prevented the…”
Distinguishing Human and AI-Generated Text: A BiLSTM-Based Deep Learning Approach
Sumaya, S. T., Arnob, H. I., Ahmed, A., Sultana, A., & Rifat, A. A. S. (2026) · 2026 IEEE 2nd International Conference on Quantum Photonics, Artificial Intelligence & Networking (QPAIN) · doi:10.1109/qpain69676.2026.11546461
“Due to the increase of the use of machine-made materials Fleckenstein et al. [18] found that teachers don’t distinguish the essays from other works and that threaten the academic institution for academic integrity.”
Who—or What—Told You? Source Memory for AI- vs. Human-Authored Content
Metzger, L., Bühler, N., Geiger, S., & Erdfelder, E. (2026) · CHI Extended Abstracts · doi:10.1145/3772363.3799350
Nyagoga, L. M., Malik, L., Lukoko, H. O., & Liu, M. (2026) · Artificial Intelligence in Education · doi:10.1108/aiie-02-2025-0029 · Cited by 2
Refinement or Replacement? Comparing Scaffolded and Traditional Prompts in AI Assisted EFL Revision
Xia, T., Liu, X., Hu, P., & Wang, J. (2026) · International Conference on Computer Science and Technologies in Education (CSTE) · doi:10.1109/cste69562.2026.11649691
“However, classroom adoption raises concerns about authenticity and over-reliance when AI revisions exceed learners' developmental level or drift from intended meaning [1],[8], [9], [15].”
Pragmatikeren, skeptikeren og entusiasten
Hanssen, G., & Leonardsen, J. K. (2026) · Skandinavisk tidsskrift for yrker og profesjoner i utvikling · doi:10.7577/sjvd.6488 · Cited by 1
“Flere studier peker på at lærere er spesielt kritiske til bruk av generativ KI eller chatbot’er i skriftlige oppgaver fordi det er utfordrende å avdekke om elevene har benyttet KI i tekstene sine (Fleckensten et al., 2024; Jeon, 2023).”
ВЫЗОВЫ ИНТЕГРАЦИИ ГЕНЕРАТИВНОГО ИИ В ПРЕПОДАВАНИЕ РУССКОГО ЯЗЫКА В КАЗАХСТАНСКИХ ВУЗАХ
Аманов, А. Ш., Клименко, И. С., & Beissembayeva, S. (2026) · Bulletin of Zhetysu University named after I Zhansugurov · doi:10.53355/zhu.2026.118.1.009
From Corpus to Cockpit: Generating Emergency Lexical Bundles in Aviation Context
Nugraha, D. S., Hasanah, E. U., Dewanti, R., & Iskandar, I. (2026) · Corpus Pragmatics · doi:10.1007/s41701-026-00234-8
“Moreover, AI supports performance-based assessment, as advocated by ICAO Doc 9835 (2010), enabling evaluation of grammatical accuracy, pragmatic appropriateness, discourse coherence, and responsiveness under pressure (Fleckenstein et al., 2024; Topping et al., 2025).”
From essay to AI-augmented writing: reframing pedagogy and feedback through the AAWP model
Askari, M., & Rahim, A. S. (2026) · Assessment & Evaluation in Higher Education · doi:10.1080/02602938.2026.2638914
Han, T., & Zheng, Y. (2026) · Studies In Educational Evaluation · doi:10.1016/j.stueduc.2026.101572 · Cited by 2
ChatGPT and Large Language Models for English Professors: Getting It Right
Sheu, C. J. (2026) · College English · doi:10.58680/ce2026883346 · Cited by 1
Review of Policies for the Use of Generative Artificial Intelligence in Leading Foreign Universities
Gorlacheva, E. N., Леонидович, О. П., & Андреевна, П. Е. (2026) · University Management Practice and Analysis · doi:10.15826/umpa.2025.04.028 · Cited by 1
Challenges Faced by ELT Teachers in Adopting ChatGPT as an Instructional Tool
Maspufah, Afriyeni, Y., Zuriati, D., & Masbiran, G. (2026) · ENGLISH JOURNAL OF INDRAGIRI · doi:10.61672/eji.v10i1.3390
“Ethical issues further complicated adoption (Baskara & Mukarto, 2023), particularly as teachers faced difficulties distinguishing between student-generated and AI-generated work (Fleckenstein & al., 2024).”
Park, S., & Han, Y.-S. (2026) · doi:10.48550/arxiv.2601.19913
“Studies compare student and LLM essays, test whether teachers and students can identify their source, and examine how source information affects ratings (Herbold et al., 2023; Waltzer et al., 2023; Fleckenstein et al., 2024).”
Vučković, A. (2025) · Edukacja Filozoficzna · doi:10.14394/edufil.2025.0006
López, G. E. C., Reyes, J. U., & Reyes, E. C. (2025) · European Public & Social Innovation Review · doi:10.31637/epsir-2026-2155 · Cited by 3
Lucas, J. (2025) · Psychology and Education A Multidisciplinary Journal · doi:10.70838/pemj.480506
Nurtanto, M., Hamid, M. A., Wahab, A. A., Kholifah, N., Gustiawan, W., & Subakti, H. (2025) · International Conference on Emerging Technologies · doi:10.1109/icet67257.2025.11291059
“Vocational teachers struggle to articulate the limits of AI use in student classwork, particularly in defining achievable goals or tasks with prohibitive clauses [3].”
OpenCoderRank: AI-Driven Technical Assessments Made Easy
Dutta, H. S., Ansari, S., Kumari, S., & Bhalerao, S. R. (2025) · arXiv.org · doi:10.48550/arxiv.2509.06774
Gorodnicheva, J., Bekim, A., Klucharev, V., Monahhova, E., Mening, S., Morozova, A., & Zinchenko, O. (2025) · 2025 Seventh International Conference Neurotechnologies and Neurointerfaces (CNN) · doi:10.1109/cnn67635.2025.11177516
“Evaluating the detectability of AI-generated texts among student essays" [11] found that teachers often could not distinguish AI-generated texts from student essays, indicating high authenticity.”
Kunstig intelligens og konsekvenser for veiledning på tekstarbeid
Kjærgård, P. I., Hodøl, H.-O., & Økland, Ø. (2024) · Nordisk tidsskrift i veiledningspedagogikk · doi:10.15845/ntvp.v9i2.4078 · Cited by 3
“Som Fleckenstein et al. sin studie viste, så kan kvaliteten på de automatisk genererte tekstene være på et slikt nivå at de kan måle seg med hva et menneske er i stand til å produsere (Fleckenstein et al., 2024).”
InfluentialDigital competence analysis to inform current pedagogy and curricula
Fuentes, A. C., & García, M. R. (2024) · Research in Education and Learning Innovation Archives · doi:10.7203/realia.33.29257
“A recent case is the need for regulating and framing GenAI resources and practices in all types of academic / educational environments, entailing a (re)formulation of DC mechanics, objectives, purposes, and implementation (e.g., see Dwivedi et al., 2023; Fleckenstein et al., 2024; Lan & Chen, 2024).”
BackgroundMetapragmatic acts performed by ChatGPT 4o and human authors in research article (RA) abstracts
Gao, Y., Chen, X., & Zhang, J. (2026) · English for Specific Purposes · doi:10.1016/j.esp.2026.05.007
Teaching AI Through Communication: Rethinking Evidence, Interaction, and Learning
Edwards, C., Craig, M. J. A., Rijhwani, V., & Abdulrauf, H. (2026) · Journal of Communication Pedagogy · doi:10.31446/jcp.2026.1.13
Metadata
- Authors
- Fleckenstein, J., Meyer, J., Jansen, T., Keller, S. D., Köller, O., & Möller, J.
- Affiliations
- University of Hildesheim (DE)
- Leibniz Institute for Science and Mathematics Education (DE)
- Zurich University of Teacher Education (CH)
- Corresponding author
- Fleckenstein, Johanna
- Venue
- Computers and Education: Artificial Intelligence
- Publisher
- Elsevier BV
- Published
- 2024-06
- Volume
- 6
- Pages
- 100209
- ISSN
- 2666-920X (ISSN-L)
- OpenAlex
- W4391222332
- Cited by
- 126
- References
- 16
- Citations by year
- 2026: 58 · 2025: 48 · 2024: 20
- FWCI
- 47.38
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Computers and Education Artificial Intelligence (publishedVersion, cc-by-nc-nd)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
- Topics
- Social Sciences › Education › Artificial Intelligence in Education
- Social Sciences › Education › Academic integrity and plagiarism
- Physical Sciences › Computer Science › Artificial Intelligence › Authorship Attribution and Profiling
- Keywords
- AI-generated text detection, generative AI, ChatGPT, student essays, teacher judgment, essay scoring, overconfidence, source effects, academic integrity
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
