In light of the disruptive advances in generative artificial intelligence (GAI), it is essential to understand how students perceive its usefulness and ease of use, as well as how these perceptions influence their attitude toward this tool and their intention to integrate it into their learning processes. In this context, having valid and reliable instruments is fundamental to assess the multiple changes GAI introduces in education. This study provides empirical evidence on the reliability of the scale Use of Generative Artificial Intelligence in Initial Teacher Education (EPIAG-FID), designed to evaluate pre-service teachers’ self-perception regarding the ethical and responsible integration of GAI as a pedagogical tool in higher education. The scale includes five dimensions: (1) Knowledge of different artificial intelligences, (2) Use of AI in their training process, (3) Ethical and responsible positioning, (4) Perceived competence in using GAI, and (5) Critical reflection on GAI's impact on education. Initially, the scale underwent expert validation with five reviewers. It was then applied to a sample of 651 pre-service teachers. The results show high reliability, both globally and by dimension. The overall Cronbach’s alpha was 0.92 and G6 was 0.94. By dimension, the lowest alpha was 0.7583 (Ethical and responsible positioning), and the highest was 0.9044 (Use of AI in training). Removing items 11 and 18 slightly improved internal consistency, though the change was marginal. All items demonstrate adequate internal consistency and contribute stably to the reliability of the scale. These findings support its future use in research and training.