Beyond the result. Didactic reflection to identify and mitigate algorithmic biases
DOI:
https://doi.org/10.31207/colloquia.v12i1.179Keywords:
Generative artificial intelligence, algorithmic bias, educational ethics, critical pedagogy, university teachingAbstract
The accelerated incorporation of generative artificial intelligence (GenAI) tools into higher education poses a series of ethical, epistemic, and pedagogical challenges that demand a critical response for and from teaching practice. This article presents a didactic reflection aimed at strengthening teachers’ skills and capacities to identify, analyze, and mitigate algorithmic biases embedded in GenAI systems applied to educational practice. Inspired by the principles of critical pedagogy and the theoretical contributions of the Frankfurt School, the proposal seeks to expose the risks of an apparent neutrality that conceals dynamics of epistemic exclusion. Through three core components—empirical analysis, critical categorization of biases, and the design of didactic strategies—the article outlines a model that can be replicated across diverse educational contexts.
To support this approach, a table is proposed to organize these components according to their main features, identification activities, and potential outcomes.
References
Adorno, T. W., & Horkheimer, M. (1998). Dialéctica de la Ilustración: Fragmentos filosóficos. Trotta.
Alvarez Newman, D. (2025). ¿Es neutral la inteligencia artificial? Dispositivo managerial tecnocrático y algoritmización del trabajo. Medellín: Biblia, Teología Y Pastoral Para América Latina Y El Caribe, 50(189), 283–308. https://revistas.celam.org/index.php/medellin/article/view/2397
Ferrante, E. (2021, julio/agosto). Inteligencia artificial y sesgos algorítmicos: ¿Por qué deberían importarnos? Nueva Sociedad, 294. https://nuso.org/articulo/inteligencia-artificial-y-sesgos-algoritmicos/
Freire, P. (2001). Política y educación (5a ed.). Siglo Veintiuno Editores.
González Padilla, A. (Ed.). (2024). Inteligencia artificial, ética y tecnofilosofía: Ensayos sobre sesgos algorítmicos, crisis ecosocial, transhumanismo y otros disruptivos. Lulaya Ediciones.
Grupo Académico de Inteligencia Artificial Generativa en Educación UNAM. (2025). Recomendaciones para el uso educativo de la Inteligencia Artificial Generativa en la UNAM (2.ª ed.). https://www.ceide.unam.mx/wp-content/uploads/2025/05/Recomendaciones_IAGEN_UNAM_2025.pdf
Ramírez, R. (2023). Sesgos y discriminaciones sociales de los algoritmos en inteligencia artificial: Una revisión documental. Revista Entretextos, 15(39). https://doi.org/10.59057/iberoleon.20075316.202339664
Runge, A. (2013). Didáctica: Una introducción panorámica y comparada [Documento no publicado]. Universidad de Antioquia. https://www.studocu.com/es-mx/document/universidad-virtual-del-estado-de-guanajuato/taller-de-lectura-y-redaccion/didactica-comparada-introduccion-panoramica-y-perspectivas-edu-201/141147254
Sued, G. E. (2022). Culturas algorítmicas: Conceptos y métodos para su estudio social. Revista Mexicana De Ciencias Políticas Y Sociales, 67(246). https://doi.org/10.22201/fcpys.2448492xe.2022.246.78422
Torres Zúñiga, V. (2025). Sesgos en la representación de profesionales de la ciencia forense en imágenes generadas por inteligencia artificial. TIES, Revista De Tecnología E Innovación En Educación Superior, 12, 26–39. https://doi.org/10.22201/dgtic.26832968e.2025.12.57
World Economic Forum. (2025). The global risks report 2025 (20th ed.). https://reports.weforum.org/docs/WEF_Global_Risks_Report_2025.pdf
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors publishing in our Journal comply to the following terms:
1. Authors keep their work's copyrights, but they guarantee Colloquia to be the first publisher of their papers. They grant the Journal with a Creative Commons Attribution License, under which their work can be shared with the condition that it is appropriately cited.
2. Authors can establish further clauses for non-exclusive distribution, such as publication on a separate book or placing in an institutional data-base. Nevertheless, a note should be always added to explain that the paper was originally published in Colloquia.
This Journal utilizes the LOCKSS system to create a file distributed among participating libraries, allowing these libraries to create permanent archives of the Journal for purposes of preservation and restoration. More information...
