JMIR Publications

ABSTRACT

Background:

Artificial intelligence (AI), health data science, digital health platforms, clinical decision support systems, and generative language models are increasingly shaping medical practice. Undergraduate medical education in Latin America requires a minimum standard that is feasible, assessable, and sensitive to regional inequalities.

Objective:

To develop an evidence-informed minimum regional AI competency-and-assessment standard for public medical schools in the expanded MERCOSUR region, grounded in a secondary curricular matrix, recent peer-reviewed literature, and international guidance on digital health, AI ethics, clinical validation, and assessment.

Methods:

We conducted a documentary evidence synthesis using a secondary curricular matrix of 131 public medical-school entries from Argentina, Brazil, Uruguay, Paraguay, and Bolivia, combined with a targeted narrative synthesis of recent peer-reviewed literature and international guidance on AI education, digital health competencies, ethics, clinical validation, and assessment. Entries without sufficient public documentation were treated as non-codifiable rather than as evidence of absence. Brazilian entries were interpreted separately because their coding was based on national curricular guidelines rather than institution-specific pedagogical projects.

Results:

Of 131 entries, 92 were codifiable and 39 were non-codifiable because of insufficient publicly available curricular documentation. Among codifiable entries, AI/bioinformatics was absent in 91 of 92 entries (98.9%), and AI/technology ethics did not exceed tangential general bioethics. We propose five minimum competencies with SMART-oriented outcomes and multimodal assessment evidence: AI and data literacy; health-data governance and digital systems; critical appraisal and clinical validation; ethics, equity, and accountability; and human-centered AI use, communication, and teamwork.

Conclusions:

The proposed framework is presented as an evidence-informed, regionally adaptable minimum standard requiring subsequent Delphi validation and pilot testing. Its main contribution is to translate documented curricular gaps into measurable outcomes that can be implemented even in low-resource settings without requiring advanced programming for every medical graduate.