From Copy-Paste to Critical Thinking: Generative AI Use in K-12 Students

Abstract

Generative Artificial Intelligence (GenAI) has transformed multiple sectors, with education being one of the most impacted. This research analyzes the use of GenAI in upper secondary students (K-12), a vulnerable and understudied sector where the central problem identifies risky practices such as uncritical use, overestimation of AI capabilities, and absence of critical thinking that generates cognitive dependence and validation of false information. This study aims to promote the development of critical thinking about GenAI use through a didactic intervention with fourteen students of 15-16 years old from the Instituto Politécnico Nacional in Mexico. Using a qualitative approach with a pretest-posttest design, the methodology employed semi-structured interviews organized in five dimensions: perception of GenAI, representation and reasoning, mediated learning, critical interaction, and ethical considerations. The results reveal significant post-intervention transformations: paradigm change from instrumental to critical-reflective use, development of verification protocols, construction of complex and structured prompts, understanding of contextual limitations, and establishment of ethical frameworks. Furthermore, this research concludes that teacher intervention can modify forms of thinking, demonstrating that with adequate pedagogical guidance, students develop competencies for responsible GenAI use, minimizing risks and maximizing benefits, these findings underscore the need for further research on teacher education, who directly influence students' perspectives and practices regarding GenAI use.