GenAI Use and Academic-Integrity Risk — Two Independent Ecuadorian University Surveys

A 2025 study in the International Journal for Educational Integrity (Torres-Diaz, Duart, Rivera and Beltran Flandoli, with lead authorship based at Universidad Tecnica Particular de Loja, UTPL) surveyed 4,811 students across eight Ecuadorian universities on generative-AI use and self-reported plagiarism. A binary logistic regression explained 20.2% of the variance in self-reported plagiarism. Two predictors stood out as counter-intuitive: students who said they disclosed their ChatGPT use to instructors, and students who trusted ChatGPT-generated information, were both more likely to self-report plagiarism than those who did not. Conversely, students who perceived AI as improving their academic performance and those with broader knowledge of AI tools reported less plagiarism. The authors note their cross-sectional design cannot establish causation, and call for ethical-awareness training, digital literacy and clearer institutional policy.

A second, independent study published in SciELO Venezuela surveyed 335 students at an Ecuadorian university on how they use ChatGPT for coursework. It found use remains limited and largely unsupervised: students report little formal guidance from instructors on appropriate use, and the measured effect on academic performance was minimal, which the authors attribute to the absence of structured guidance and ethical framing around the tool.

Read together, the two surveys — run by different teams, at different scales, with different instruments — point to the same gap: Ecuadorian universities have widespread GenAI use among students but little institutional scaffolding around it. Neither study evaluates a specific training or policy intervention, so this entry is included as evidence of risk and unmet need rather than a proven solution.

Read the full analysis: https://ve.scielo.org/scielo.php?script=sci_arttext&pid=S2665-02822025000100085