Brazil's ENAMED and the Governance of AI-Supported Exam Preparation: A Conceptual Policy Analysis
Year: 2026
Type: article
Abstract: Brazil's National Examination for the Assessment of Medical Training (ENAMED) creates a policy setting in which a national assessment can influence student preparation, institutional priorities, and access to residency selection. At the same time, generative artificial intelligence and adaptive learning tools are becoming readily available to medical students. This conceptual policy analysis asks how medical schools should govern AI-supported preparation before high-stakes incentives define its educational purpose. We use a targeted narrative analysis of official ENAMED documents and selected literature on high-stakes testing, medical assessment, artificial intelligence in medical education, and faculty development. We argue that personalisation can become adaptive test coaching when systems optimise visible examination performance without evidence of transfer to clinical reasoning, feedback literacy, or professional judgement. Governance should therefore address educational purpose, evidence standards, transparency, human oversight, equity, data protection, conflicts of interest, and continuous monitoring. Medical schools should align AI-supported preparation with programmatic assessment, prepare faculty and learners to evaluate outputs critically, and keep consequential decisions under accountable human control. ENAMED offers an opportunity to establish these safeguards early. The appropriate question is not whether students will use AI for preparation, but whether institutions will shape that use toward learning and public accountability rather than narrow score optimisation.
Language: English
Subfield: Health Informatics
Field: Medicine
Domain: Health Sciences
Open Access status: hybrid
OpenAlex ID: W7203465582