Motivation, Ethics, and Trust: A Data-Informed Analysis of Generative AI's Impact on EFL Students

Abstract

This mixed methods study provides a data-informed analysis of the impact of generative artificial intelligence (GenAI) on the motivation, English proficiency, and ethical perceptions of English as a foreign language students. Grounded in self-determination theory and the unified theory of acceptance and use of technology 2, the study employed a pretest/posttest control group design (N = 310) with semi-structured interviews (n = 25). The GenAI intervention induced statistically significant improvements in student motivation and English proficiency compared to the control group. A key finding from a regression analysis (n = 155) was that GenAI use frequency was the sole significant predictor of motivation (β = .46, p < .001), superseding preexisting proficiency or anxiety and suggesting a democratizing effect. Furthermore, perceptions of the tool's transparency and fairness were strong, direct predictors of motivation. Counterintuitively, concerns about bias were positively correlated with motivation and trust, indicating that practical use fosters a more critical AI literacy rather than naive acceptance. Qualitative data confirmed that this newfound critical awareness led students to advocate for a balanced, human-in-the-loop pedagogical model. The findings demonstrate that successful GenAI integration requires a dual focus on motivational enhancement and fostering critical, ethical engagement, with the human educator positioned as an essential guide.