GENERATIVE ARTIFICIAL INTELLIGENCE AND EDUCATIONAL ASSESSMENT: A DOCUMENTARY ANALYSIS

Abstract

Discussions about the accelerated development of artificial intelligence and its uses in various sectors of society have increased significantly. In education, in the last year, we have gone from total rejection to dissemination of research about its benefits for learning. With these aspects in mind, this study investigates Brazilian and foreign public documents published between 2022 and 2025 that present guidelines for the use of AI in education and, more specifically, understands how they approach assessment, also seeking information on the use of AI in assessment for and in the context of language teaching. Thus, we answer the following questions: what do Brazilian and international documents on AI in education propose in relation to assessment? How can these guidelines impact assessment in language teaching? To this end, we conducted qualitative, bibliographical and documentary research, and used content analysis as a technique to categorize the data found. The results point to formative assessments through task design to assess skills, competencies and performances through the use of AI. Furthermore, they warn about the critical use of AI in educational contexts, aiming at the development of some aspects, such as ethics in academic production, the mediating role of teachers in the classroom, creativity and human qualification. From this perspective, we highlight the need to promote literacy in AI and conclude by emphasizing the urgency of assessments having a formative focus, taking advantage of the potential of AI tools but not relying solely on them.