Bridging minds and machines: perception of artificial intelligence in scientific writing across Latin America

This study examined the acceptance of artificial intelligence (AI) tools for scientific writing among 71 graduate students from sampled institutions in Mexico, Ecuador, and Colombia who used AI during a three-month workshop focused on the development of a research protocol. Data were collected through a questionnaire conceptually informed by the UTAUT2 model and a rubric designed to assess students' performance in dimensions such as writing efficiency, organization of ideas, reference management, and autonomy. The results showed generally favorable perceptions of AI, particularly regarding its usefulness for tasks related to literature review and language precision. However, the rubric-based evaluation revealed more moderate performance, especially in autonomy and skills development, suggesting a difference between students' positive perceptions of AI and their actual use of these tools in academic writing tasks. Correlational analyses indicated positive associations among perceived usefulness, ease of use, and perceived value, while descriptive and inferential comparisons showed some variation across the sampled country groups. Overall, the findings provide an exploratory view of how AI is perceived and used in graduate-level scientific writing contexts and highlight the importance of promoting autonomous, critical, and ethically informed use of these tools in higher education.