AI Competency Framework and a Proposal for Teacher Training

Luzia Bueno; Alaina Alves

DOI: 10.22167/2675-6528-202603217

Article derived from a Course Conclusion Work (TCC), with content based on the student’s original work and adapted to the editorial format of the E&S Magazine with the support of the ResumeAI tool, an artificial intelligence solution developed by Instituto Pecege for textual synthesis and organization.

Summary

The articulation between neuroscience and education for human development, allied to the relevance of prescriptive texts in the training of teachers in Brazilian higher education, motivated the investigation. The study aimed to investigate the potential of UNESCO’s Framework of Competencies in AI for Teachers (MRCP-IA) document for teacher training that prioritizes human development and, from this analysis, to present a proposal for its implementation in the Brazilian context, articulated with the principles of neuroscience. The research was characterized as descriptive, qualitative, and documentary analysis. For the analysis of the MRCP-IA, the theoretical-methodological framework of Sociodiscursive Interactionism was employed. The interpretation of the results was based on discussions from the Sciences of Work, Sociodiscursive Interactionism, Vygotsky, and the 12 principles of neuroscience listed by Amaral and Guerra (2022). The results indicated that the MRCP-IA positions Artificial Intelligence as the protagonist, assigning to the teacher the responsibility of dealing with it ethically and critically, but without detailing teacher training. Therefore, a proposal for teacher training for Pedagogy courses was formulated, considering the principles of neuroscience and Vygotsky’s historical-cultural perspective, emphasizing social interaction and the integral engagement of the teacher. It was concluded that continuing education, which respects brain individuality and promotes social interaction with the teacher’s bodily and emotional engagement, can effectively drive human development.

Keywords: Competencies; AI; Neuroscience; Pedagogy; Teaching work.

1. Introduction

Neuroscience, an interdisciplinary field that investigates the nervous system, has become increasingly relevant to education. According to Amaral and Guerra (2022), learning directly impacts the development of the nervous system. Thus, the integration of neuroscience advances into educational practices constitutes a crucial trend for optimizing the development of students and teachers, by providing greater clarity on the necessary investments to enhance higher mental functions.

In this scenario, the reflection on how the various artifacts available to the teacher, from higher education to basic education, can mediate the dialogue between neuroscience and education is fundamental. Among these artifacts, prescriptive documents play a central role, guiding the teacher’s work at all levels of education. Such documents, which form a “cascade of prescriptions” (Abreu-Tardelli, 2006), originate from international organizations such as UNESCO and OECD, reaching national and local guidelines (Bronckart and Machado, 2004; Bueno et al., 2023).

The understanding and analysis of these documents are essential for effective intervention in teacher training processes, aiming to improve the quality of Brazilian education. Recently, UNESCO’s Framework of Artificial Intelligence Competencies for Teachers (AI-FCTC) has emerged as a document of great importance, as it seeks to guide educators on the appropriate use of Generative Artificial Intelligence (AI) in education. This framework is particularly relevant, considering that, as of 2022, few countries had established similar guidelines for their educators.

However, the rapid rise of AI in education, while promising to facilitate tasks, demands critical reflection on its impact on human development. The delegation of capabilities such as selecting, reading, reflecting, and analyzing information can affect higher mental functions, contradicting Vygotsky’s perspective (2007; 2009) on the importance of social interaction and active human agency. Furthermore, MRCP-AI does not directly address emotion as a fundamental component of development, an aspect central to Vygotskian theory.

The appropriation of artifacts, including AI, is crucial for them to become psychological instruments that drive teacher development, rather than generating illness (Clot, 2008, 2010; Machado, 2007). However, the UNESCO document, by focusing on AI as the protagonist and assigning the teacher the responsibility of acquiring competencies, does not detail how this training can be effectively implemented in the Brazilian context. Recent data reveal significant challenges in teacher training in Brazil, such as the low evaluation of Pedagogy courses and the expansion of Distance Learning with quality problems (Meneghel, Sousa Filho e Ferraz, 2024; Nascimento, Silva e Moreira, 2024).

Additionally, the MRCP-IA, being a global document, does not consider Brazil’s regional disparities and technological infrastructure, which may hinder the implementation of training that requires constant access to the internet and digital tools. The enormous responsibility attributed to the teacher in building a better society with AI, in a scenario of precariousness and platformization of education (ABMES, 2023; Oliveira, 2024; Matiola, Santos e Neto, 2023; Santos, 2025), raises questions about the real capacity of teachers to reverse market decisions and guarantee human autonomy in the face of technology.

Given this panorama, it becomes imperative to develop a teacher training proposal that not only aligns with the principles of neuroscience but also considers the specificities of the Brazilian educational context. Such a proposal should aim at the critical appropriation of AI as an instrument for human development, promoting social interaction, respect for the singularities of each brain, and the mobilization of emotions in the learning process. The relevance of the articulation between neuroscience and education for human development, added to the importance of prescription texts in teacher training, justifies the need to investigate the potential of the UNESCO document “Reference Framework of Artificial Intelligence Competencies for Teachers” (RAIC-AI) to contribute to a training where human development is the priority and, from this analysis, to present a proposal for the implementation of the RAIC-AI in the Brazilian context, articulated with the principles of neuroscience.

2. Material and Methods

The research conducted was characterized as descriptive and qualitative, with a documentary analysis approach. The central methodological objective was to investigate the potential of the UNESCO Framework of Competences in AI for Teachers (MRCP-IA), aiming to support a teacher training proposal in the Brazilian context, articulated with the principles of neuroscience, according to the general objective of the study.

To this end, the UNESCO MRCP-IA document was selected as the main empirical object. This 58-page document is publicly available on the UNESCO website, accessed on October 19, 2025, via the link https://unesdoc.unesco.org/ark:/48223/pf0000394280/PDF/394280por.pdf.multi. Its choice is justified by its global relevance and the absence of similar regulatory frameworks in many countries until 2022, as highlighted in the introduction.

The analysis of MRCP-IA was conducted under the theoretical-methodological framework of Sociodiscursive Interactionism (ISD), as proposed by Bronckart (1999, 2006, 2022). This approach, based on Vygotsky’s studies (2007, 2009), allows us to understand how language in texts contributes to the morphogenesis of human action, revealing implicit or explicit models of action. ISD was employed to unveil the figures of action present in the document, identifying the highlighted dimensions, such as protagonists, actions, and purposes.

The application of ISD followed a sequence of analytical steps to examine the document. Initially, the context of production of the MRCP-IA was analyzed, considering its origin and authorship. Then, the textual architecture was analyzed, which included the textual infrastructure (global plan of thematic content, discourse types, and sequence types), the textualization mechanisms (connection, nominal cohesion, and verbal cohesion), and the enunciative mechanisms (modalization and voices).

Complementarily, the protagonists of the document were identified, seeking to determine who the central agents of the proposed actions are. Subsequently, the actions attributed to these protagonists were detailed, focusing on verbs and expressions that indicate obligations or possibilities. To aid in identifying patterns and term frequency, such as the occurrence of keywords, text editing tools were used for simple word counting.

To ground the analysis and reflection on the document’s implications for teaching work and human development, discussions from the Sciences of Work were employed, according to Amigues (2004) and Clot (2010, 2017). Additionally, the perspective of Sociodiscursive Interactionism and the studies of Vygotsky (2007, 2009) were mobilized to deepen the understanding of appropriation and development processes.

The 12 principles of neuroscience, listed by Amaral and Guerra (2022, p. 91-93), also integrated the theoretical framework for the analysis. These principles were considered to evaluate how didactic proposals, articulated with the MRCP-IA, could stimulate brain activities and promote human development, respecting individuality and social interaction.

Based on the analysis of the document and the theoretical foundation, we sought to reflect on the potential of MRCP-IA for teacher training that prioritizes human development. This reflection served as the basis for the elaboration of a proposal for teacher training for Pedagogy courses, centered on human interaction, on real and proximal development zones, and on the development of executive functions.

3. Results and Discussion

The analysis of UNESCO’s AI Competency Framework for Teachers (A I C F – T) revealed that the document, with its 58 pages, was developed by renowned experts in technology and education, lending it considerable legitimacy. The text is presented in the third person, with verbs in the present tense, and adopts a predominantly expository and prescriptive character, suggesting a consensus on the need for teachers to adapt to the reality of Artificial Intelligence. The central purpose of the A I C F – T is to fill the regulation gap on AI in most countries, serving as a basis for future guidelines and teacher training, and guiding teachers towards a new educational era in which the teacher-student relationship extends to teacher-AI-student. To this end, the document defines essential knowledge, skills, and values for teachers in the age of AI.

The MRCP-IA structures competencies into five dimensions and three progression levels. The dimensions include “human-centered mindset”, “AI ethics”, “AI fundamentals and applications”, “AI pedagogy”, and “AI for professional development”. The progression levels are “Acquire”, “Deepen”, and “Create”, indicating an evolution from basic understanding to innovation. However, a simple word count of the document revealed that Artificial Intelligence is the central protagonist, mentioned 2638 times, while the term “teacher” or its variations appear only 547 times. This disparity suggests a greater focus on the technology than on the human agent who will use it.

The document highlights both the positive aspects of AI, such as its potential for information processing, data analysis, content generation, and the use of multimodality, as well as the associated risks. Among the concerns raised are the threat to individual autonomy, environmental impact, privacy violation, and the exacerbation of social disparities and prejudices. In the pedagogical context, the MRCP-AI warns of the danger of automating education, underestimating teacher mediation, and weakening the teacher-student bond, in addition to diverting resources to technology instead of valuing education professionals.

The figure of the professor is constructed in the MRCP-IA as someone who has a set of obligations towards AI. The frequent use of deontic modal verbs, such as “must” (28 times) and “needs” (23 times), reinforces the normative and imperative nature of the actions expected of teachers. For example, the document states that “Teachers must remain responsible for pedagogical decisions in the use of AI in teaching and in facilitating its use by students” (UNESCO, 2025, p. 17). This language suggests that the responsibility for the ethical and effective integration of AI falls primarily on the professor, without detailing how this training can be effectively implemented.

At the “Acquire” level, the expected competencies for teachers involve maintaining human control over technology, protecting rights and diversity, understanding the basic functioning of AI (data and algorithms), planning and teaching while mitigating risks, and personalizing their professional development. For the “Deepen” level, educators are expected to question exaggerated promises of AI, protect data and copyrights, master specific algorithms and tools, personalize teaching for each student, and collaborate in teacher networks through AI. At the “Create” level, competencies include understanding the impact of AI to build fair and inclusive technology, following and leading the creation of ethical rules, operating and creating with AI, evaluating and innovating in teaching, and validating professional and community development strategies.

Despite the breadth of these competencies, the MRCP-IA does not directly refer to emotion as a fundamental component of human development. This omission is notable, especially when returning to Vygotsky’s perspective (2007; 2009), which emphasizes emotion as an integral part of the development process. The absence of this aspect may limit the understanding of the appropriation of AI as a psychological instrument that mobilizes both cognitive and affective issues, essential for the health of teaching work, as discussed by Clot (2008, 2010).

Another limitation identified is the linear conception of skills progression (“Acquire – Deepen – Create”). Vygotsky’s (2007; 2009) theory on human development points to a non-linear process, with leaps and “setbacks,” which vary for each individual. Therefore, training for groups that does not consider these individual differences may not be effective, as the heterogeneity of development levels within the same group, if well managed, can drive the development of all participants.

Furthermore, the global nature of the MRCP-IA does not consider the specificities of the Brazilian context, such as regional disparities and technological infrastructure. The implementation of training that requires constant access to quality internet and digital tools can be a significant challenge in many regions of the country. This lack of contextualization can compromise the applicability of the document and the effectiveness of the proposed training, distancing it from the reality experienced by Brazilian teachers.

The document assigns an enormous responsibility to the professor in building a better society with AI, requiring educators to understand the technology, be critical, and care for its development and the community. However, this expectation clashes with the reality of precariousness and platformization of teaching in Brazil. Data from the Higher Education Census (ABMES, 2023) indicate that Distance Education (EaD) grew by 700% in ten years, with an average of 171 students per professor in the private sector. In some institutions, this proportion is even higher, reaching 2,594 students per professor or 1,325 students per professor, with a 26.7% reduction in the teaching staff (Oliveira, 2024).

Census data (BRASIL, 2025) reinforce this trend, showing that between 2014 and 2024, the total number of higher education faculty decreased from 383,386 to 374,501, while enrollment increased from 7 million to over 10 million. This scenario of overload and precariousness of teaching work (Matiola, Santos, and Neto, 2023; Santos, 2025) raises doubts about the real capacity of professors to assume the responsibility of reversing market decisions and guaranteeing human autonomy in the face of technology, as MRCP-AI seems to suggest.

Teacher education in Brazil, especially in Pedagogy courses, faces significant challenges. Evaluations from the National Student Performance Exam (ENADE) indicate that only 25% of Pedagogy students evaluated achieve scores of 4 and 5, on a scale of 0 to 5 (Meneghel, Sousa Filho e Ferraz, 2024; Nascimento, Silva e Moreira, 2024). Furthermore, the expansion of Distance Learning (EaD) has resulted in over 60% of graduates coming from this modality, with many institutions showing low evaluations. This weakened context makes it difficult to implement training consistent with the expectations of MRCP-IA, which presupposes continuous and in-depth development of competencies.

Given these considerations, a teacher training proposal for Pedagogy courses was formulated, articulated with the principles of neuroscience and Vygotsky’s historical-cultural perspective (2007; 2009). This proposal aims at a critical appropriation of AI as an instrument for human development and is organized into three stages. The first, called “AI Tools Appropriation”, focuses on the experience and exploration of the tools, with repetition and reflection on their use, consolidating knowledge. It includes critical readings, production of oral and written texts, and favors peer interaction, with attention to the Zone of Proximal Development (ZPD).

The second phase, “Deepen,” encourages future teachers to compare tools for the same objective and to seek new ones in other contexts. Reading, listening, and text production activities are maintained, but with more complex genres, such as oral presentations at events and scientific articles. Interaction and ZPD continue to be central elements in this stage, promoting deeper engagement with technology and its pedagogical implications. This phase aims to expand teachers’ critical and analytical capacity.

The third moment, “Create,” considers the professor’s desire to innovate, leveraging the previous stages to review tools and propose modifications. Participants are encouraged to share their creations with colleagues through workshops, seminars, and the production of scientific articles. The issue of emotions is explicitly considered, recognizing that knowledge and development are also driven by affective engagement, in line with Vygotskian theory. Social interaction and the ZPD remain fundamental to the success of this stage.

This training proposal mobilizes the 12 principles of neuroscience by Amaral and Guerra (2022). Principle 1 (Learning modifies the brain), Principle 6 (Motivation puts the brain into action), and Principle 7 (Attention is the gateway to learning) are activated by pedagogical strategies that promote the reorganization of brain connections. Principle 2 (The way each person learns is unique) and Principle 3 (Social interaction favors learning) are contemplated by considering each person’s real developmental level in groups, ensuring everyone’s progress.

The critical view of AI and the focus on executive functions (planning, selecting, inhibiting, and flexing actions) address Principle 4 (The use of technology influences information processing and storage) and Principle 8 (The brain is not multitasking). The teacher’s protagonism, the ability to make choices and decisions, and the mobilization of emotions in the learning process align with Principle 5 (Emotion guides learning) and Principle 10 (Self-regulation and metacognition enhance learning). Creative actions exercise the brain, according to Principle 12 (Creativity reorganizes multiple brain connections).

The training also considers the need for time for brain development, with processes of repetition, elaboration, recall, and consolidation, in accordance with Principle 9 (Active learning requires elaboration and time for consolidation in memory). Finally, the inclusion of activities that mobilize the body and different senses aims to make learning more effective, in line with Principle 11 (When the body participates, learning is more effective). Thus, AI, from a mere artifact, can be transformed into a psychological instrument that drives the development of the teacher, their colleagues, and students.

In summary, the research revealed that MRCP-IA, despite being a relevant document for guiding the integration of AI in education, presents significant limitations by positioning AI as the protagonist, neglecting the emotional dimension of human development, proposing a linear progression of skills, and disregarding the specificities of the Brazilian context and the precariousness of teaching work. The three-moment training proposal, articulated with the principles of neuroscience and Vygotsky’s historical-cultural perspective, seeks to overcome these gaps, promoting a critical appropriation of AI that values social interaction, integral engagement, and respect for the teacher’s individual cerebral specificities, effectively aiming at human development.

4. Conclusion

The research investigated the potential of UNESCO’s Framework of Artificial Intelligence Competencies for Teachers (AI-FCTC) for teacher training that prioritizes human development, seeking, from this analysis, to present a proposal for its implementation in the Brazilian context, articulated with the principles of neuroscience. It was found that the AI-FCTC, although a document of great legitimacy and scope, positions Artificial Intelligence as the central protagonist, assigning to the teacher the responsibility of dealing with it ethically and critically, but without detailing how this training could be effectively implemented. It was identified that the document neglects the emotional dimension of human development, proposes a linear progression of competencies that does not align with the non-linear nature of individual development, and disregards the specificities of the Brazilian educational context, including regional disparities and the precariousness of teaching work. It was observed that the overload and low evaluation of Pedagogy courses in Brazil hinder the insertion of continuous and in-depth training, such as that suggested by the AI-FCTC.

Faced with these gaps, a teacher training proposal for Pedagogy courses was formulated, articulated with the 12 principles of neuroscience by Amaral and Guerra (2022) and the historical-cultural perspective of Vygotsky (2007; 2009). This proposal, organized into three moments — Appropriation of AI tools, Deepen, and Create —, aims for a critical appropriation of Artificial Intelligence as an instrument for human development. The training seeks to overcome the linear view of competencies by considering brain individuality and the importance of social interaction, promoting the teacher’s bodily and emotional engagement. By integrating emotion as a fundamental component and respecting different developmental rhythms, the proposal contributes to AI becoming a psychological instrument that drives the development of the teacher, their colleagues, and students, strengthening higher mental functions and the health of teaching work.

Bibliographic References

ABMES. 2023. Associação Brasileira De Mantenedoras De Ensino Superior. Censo da Educação Superior aponta que EAD cresceu 700% em 10 anos; média é de 171 alunos por professor na rede privada. Brasília, DF, 11 out. 2023. Disponível em: https://abmes.org.br/noticias/detalhe/4968/censo-da-educacao-superior-aponta-que-ead-cresceu-700-em-10-anos-media-e-de-171-alunos-por-professor-na-rede-privada. Acesso em: 19 jan. 2026.

ABREU-TARDELLI, L. S. 2006. A cascata de prescrições no trabalho docente: o processo de elaboração de documentos oficiais em foco. Trabalhos em Linguística Aplicada, Campinas, v. 45, n. 2, p. 195-212.

AMARAL, A. L. N.; GUERRA, L. B. 2022. Neurociência e educação: olhando para o futuro da aprendizagem. Brasília: SESI/DN. 290 p.

AMIGUES, R. 2004. Trabalho e ofício do professor: as competências de ofício. Lisboa: Educa.

BRASIL. 2025. Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira (Inep). Censo da Educação Superior 2024: notas estatísticas. Brasília, DF: Inep. Disponível em: https://download.inep.gov.br/publicacoes/institucionais/estatisticas_e_indicadores/notas_estatisticas_censo_escolar_da_educacao_superior_2024.pdf. Acesso em: 19 jan. 2026.

BRONCKART, J.-P. 1999. Atividades de linguagem, textos e discursos: por um interacionismo sociodiscursivo. Tradução de Anna Rachel Machado e Péricles Cunha. São Paulo: EDUC.

BRONCKART, J.-P. 2006. O agir dos discursos: dos conceitos da psicologia do desenvolvimento aos instrumentos de análise de práticas sociais. Campinas: Mercado de Letras.

BRONCKART, J.-P. 2021. Teorias da linguagem: nova introdução crítica. Tradução de Luzia Bueno, Ana Maria Mattos Guimarães e Eliane Gouvêa Lousada. Campinas, SP: Mercado de Letras.

BRONCKART, J.-P.; MACHADO, A. R. 2004. Procedimentos de análise de textos sobre o trabalho educacional. In: MACHADO, A. R. (Org.). O ensino como trabalho: uma abordagem discursiva. Londrina: Eduel. p. 131-163.

BUENO, L. et al. 2023. Dimensões da oralidade em documentos oficiais da Educação no Brasil. Palavras, Lisboa, Portugal: Associação de Professores de Português, n. 60-61, p. 129-152.

CLOT, Y. 2010. Trabalho e poder de agir. Belo Horizonte: Autêntica.

CLOT, Y. 2017. A função psicológica do trabalho. Petrópolis: Vozes.

MACHADO, A. R. 2007. Por uma concepção ampliada do trabalho do professor. In: GUIMARÃES, A. M. M.; MACHADO, A. R.; COUTINHO, A. (Org.). O interacionismo sociodiscursivo: questões epistemológicas e metodológicas. Campinas: Mercado de Letras. p. 77-97.

MATIOLA, R. P.; SANTOS, R.; NETO, A. S. 2023. Dos usos das tecnologias ao fenômeno de Plataformização da Educação: uma perspectiva crítica de análise do trabalho docente. Educação e Fronteiras, Dourados, MS, v. 13, n. 37, p. 119-136. Disponível em: https://ojs.ufgd.edu.br/index.php/educacao/article/view/16782. Acesso em: 19 jan. 2026.

MENEGHEL, S. M.; SOUSA FILHO, E. M.; FERRAZ, F. S. 2024. Perfil E Desempenho De Cursos De Licenciatura No Brasil Em Uma Década De Enade (2011-2021). PESQUISA EM FOCO, v. 29, n. 1. Disponível em: https://ppg.revistas.uema.br/index.php/PESQUISA_EM_FOCO/article/view/3741. Acesso em: 19 jan. 2026.

NASCIMENTO, C. P.; SILVA, F. T.; MOREIRA, J. R. 2024. Os cursos de pedagogia no Brasil: uma análise quali quantitativa do perfil das instituições a partir dos resultados do ENADE – 2021. Caderno Pedagógico, v. 21, n. 10, p. e8788. Disponível em: https://ojs.studiespublicacoes.com.br/ojs/index.php/cadped/article/view/8788. Acesso em: 19 jan. 2026.

OLIVEIRA, C. 2024. MEC identifica faculdades de EaD com mais de 500 alunos por professor. Extra Classe, Porto Alegre, 18 jan. 2024. Disponível em: https://www.extraclasse.org.br/educacao/2024/01/mec-identifica-faculdades-de-ead-com-mais-de-500-alunos-por-professor/. Acesso em: 19 jan. 2026.

SANTOS, D. G. 2025. Plataformização Da Educação Pública Paranaense: Reflexões Sobre Autonomia Docente, Adoecimento Controle E Vigilância. Revista Eletrônica Multidisciplinar de Investigação Científica, Brasil, v. 4, n. 22. Disponível em: https://remici.com.br/index.php/revista/article/view/765. Acesso em: 19 jan. 2026.

UNESCO. [s.d.]. Marco referencial de competências em IA para professores. Paris: UNESCO. Disponível em: https://unesdoc.unesco.org/ark:/48223/pf0000394280/PDF/394280por.pdf.multi. Acesso em: 19 out. 2025.

VYGOTSKY, L. S. 2007. A formação social da mente: o desenvolvimento dos processos psicológicos superiores. Tradução de José Cipolla Neto et al. 7. ed. São Paulo: Martins Fontes.

VYGOTSKY, L. S. 2009. A construção do pensamento e da linguagem. Tradução de Paulo Bezerra. 2. ed. São Paulo: Martins Fontes.

Article originating from a Final Coursework on Neuroscience and Learning in Education

To learn more about the course, click here and access the MBX Academy platform