Effect of generative artificial intelligence training on the competencies of border guard agents.

Abstract

This study assessed whether a generative artificial intelligence training model develops the professional competencies of border guard agents in training more than traditional instruction. A quasi-experimental pretest and posttest design was applied to 168 agents at a training school of the Panamanian border police, assigned by platoon either to an educational technology platform with generative feedback or to conventional instruction for eight weeks. Both groups improved significantly, with the largest gain in technological competence, but the group by time interaction was not significant in any dimension. We conclude that generative artificial intelligence did not outperform intensive professional training, although small effects cannot be ruled out, which points to its use as a scalable complement.