Journal of
eISSN: 2573-2897
Submit manuscript...
Research Article Volume 11 Issue 2
Artificial intelligence training and population bias: underrepresentation of Latin America in gender, bioethical and clinical decision algorithms
Pedro Hecht-López
Received: April 15, 2026 | Published: May 28, 2026
Citation: Hecht-López P. Artificial intelligence training and population bias: underrepresentation of Latin America in gender, bioethical and clinical decision algorithms. J His Arch & Anthropol Sci. 2026;11(2):90-95. DOI: 10.15406/jhaas.2026.11.00344
Abstract
Background: Artificial intelligence systems are increasingly integrated into clinical and decision-making processes, yet their performance is strongly influenced by the representativeness of training data. Concerns have emerged regarding algorithmic bias affecting gender, ethnicity, and geographic populations, particularly in underrepresented regions such as Latin America. Methods: A mini-review was conducted using the Google Scholar database to identify review articles published between 2024 and 2026 addressing artificial intelligence, bias, clinical applications, and population representation. A total of 2,820 records were identified, from which the 20 most relevant articles were screened according to predefined search terms and eligibility criteria. Following exclusion procedures, 18 review articles were included and categorized based on the presence and combination of search terms in their titles. Results: No articles simultaneously included all predefined search terms. The most frequent category involved studies addressing artificial intelligence and bias, whereas fewer articles incorporated clinical or algorithmic dimensions. Only a limited proportion explicitly referenced Latin America, and none integrated this regional focus with bias, gender, and clinical decision-making variables. The distribution of terms revealed a fragmented literature landscape with limited multidimensional integration. Conclusion: The findings suggest the existence of an important gap in the current literature, indicating that the limited representation of Latin American populations may contribute to persistent gender, bioethical, and clinical biases within artificial intelligence decision algorithms. These results highlight the importance of developing more inclusive and context-sensitive research frameworks.
Keywords: artificial intelligence, algorithmic bias, Latin American population, clinical decision algorithms, bioethics, gender bias, data underrepresentation, healthcare equity
©2026 Hecht-López. This is an open access article distributed under the terms of the, which permits unrestricted use, distribution, and build upon your work non-commercially.