Despite its potential to personalize learning and automate tasks, the adoption of artificial intelligence (AI) in school settings has been uneven, due to training, ethical, and technological limitations. This study examined the level of AI knowledge among high school teachers at Universidad Juárez in Durango, Mexico, aiming to identify strengths and gaps in their technological literacy. A quantitative, descriptive approach was employed, surveying a sample of 72 teachers using a questionnaire validated by experts. The study assessed dimensions such as comprehension, familiarity, interaction, ethical implications, and overall knowledge level. Results showed that 58.3% of teachers demonstrated a moderate understanding of AI, while only 20.8% reached a high level. Familiarity was low in 50% of participants, with just 5.6% showing high familiarity. Additionally, 38.9% reported limited ability to interact with AI systems, and the same percentage failed to identify ethical implications of AI use. Only 20.8% demonstrated a high level of comprehensive knowledge. These findings align with previous studies highlighting fragmented literacy and limited critical training. The study concluded that, while there are emerging advances, the use of AI in teaching remains constrained by structural gaps. It recommends strengthening training programs that address not only technical use, but also ethical thinking and pedagogical reflection, to prevent a superficial or uncritical integration of these technologies in the classroom.

Keywords: Artificial intelligence, teachers, digital literacy, educational ethics, education.

Artificial intelligence (AI) has established itself as a key technology in the transformation of strategic sectors such as healthcare, industry, the economy, and increasingly, education. Its capacity to adapt learning experiences, streamline administrative tasks, and provide real-time feedback is reshaping conventional teaching methods (Luckin, 2018; García Peñalvo et al., 2024). In this context, the role of educators is undergoing significant changes, requiring not only the development of essential digital skills but also a thoughtful and ethical understanding of how these new technologies work, their potential, and their limitations.
On the other hand, García Peñalvo et al. (2024) argue that the integration of AI poses considerable challenges for educator development, as inadequate specialized training limits the effective implementation of these technologies. Despite AI's clear benefits in education, the general lack of foundational knowledge and practical applications continues to hinder its full potential in the educational field. As noted by Martinell and Alvarado (2024), many educators still lack adequate comprehension of AI and its articulation with pedagogical practice.

The need for educators to develop a solid understanding of AI’s usefulness and applications in teaching is not merely a technical requirement but a substantive imperative to enable more inclusive, efficient, and equitable educational experiences. However, the meaningful integration of AI in teaching practices depends largely on educators' advanced digital competencies, their openness to adopting novel tools, and the institutional frameworks supporting learning management systems and personalized assistance technologies that enable the acquisition of the necessary skills for AI implementation and use (Russell & Norvig, 2021; Russell & Norvig, 2004).

Numerous studies concur that educators demonstrate varying levels of understanding and familiarity with AI, a phenomenon influenced by factors such as academic qualifications, professional experience, the nature of their institution, and access to ongoing professional development networks (Martinell & Alvarado, 2024; Chandio et al., 2022). This disparity leads to what is known as “fragmented learning,” where educators may grasp AI’s basic concepts and functionalities but lack advanced technical skills and, more importantly, a critical perspective on its pedagogical application (Ding et al., 2024; Zhang & Zhang, 2024).

Moreover, current academic discourse increasingly emphasizes the importance of human-centered design as a methodology for improving interactions between educators and AI systems. This approach focuses on user experience, accessibility, and usability as core criteria for developing educational technologies (Wongso et al., 2024). Similarly, continuous interaction models have been proposed as alternatives to static and inflexible systems, offering educators greater autonomy to adapt and co-develop technological solutions (Wintersberger et al., 2022).

Regarding ethics, there is a notable lack of awareness about the ethical ramifications of AI implementation in educational contexts. Most educators fail to adequately recognize the dilemmas posed by the deployment of intelligent systems, including issues related to student data privacy, algorithmic bias, automated decision-making, and surveillance without explicit consent. This oversight threatens fundamental principles of educational equity and justice and underscores the need to embed ethical frameworks within pedagogical curricula (Al-Adwan, 2025).

This study, conducted at the high school of Universidad Juárez in Durango, Mexico, confirms the presence of fragmented and inconsistent knowledge about AI among educators. While some initial progress has been made in conceptual understanding, familiarity, interaction skills, and ethical awareness remain insufficient to support a critical and meaningful integration of these technologies. This educational gap not only hampers the pedagogical potential of AI but also perpetuates a culture of technological trust devoid of comprehension, where automated decisions may substitute the judgment of ill-prepared educators.

In the context of teacher training, this scenario represents a fundamental structural issue. The lack of specific training in AI, limited promotion of collaborative frameworks for its application, and the absence of institutional strategies to enhance advanced digital literacy among teachers highlight the urgent need to re-evaluate professional development models for educators. If these deficiencies are not addressed, there is a significant risk that AI will be used uncritically, in a merely functional or operational way, thereby undermining the principles of quality, inclusion, and educational equity that should underpin any technological advancement.

The general objective of this study was to assess the level of AI knowledge among high school teachers at Universidad Juárez in the state of Durango, Mexico, in order to identify strengths, limitations, and potential training needs for the educational use of this technology.

Type of Research

This research followed a quantitative, descriptive, and straightforward approach aimed at understanding in order to act, build, or modify specific realities (Sánchez & Reyes, 2006, p. 37). Likewise, a descriptive design was adopted, whose main objective was to describe some fundamental characteristics of the studied population, based on systematic criteria that make it possible to reveal the structure or behavior of the problem addressed.

Population and Sample

The study population consisted of 100 teachers, of whom 44 were women and 56 men. Among them, 92 were full-time faculty and 8 were substitutes. A random sample of 72 teachers was selected, with 51.4% male (n=37) and 48.6% female (n=35), indicating a relatively balanced gender distribution. In terms of age group, the majority of participants were classified as adult educators, representing 54.2% of the total (n=39). Regarding the semester in which they taught, the data showed a relatively even distribution: 33.3% of the teachers (n=24) taught in the first semester, 31.9% (n=23) in the second semester, and 34.8% (n=25) in the third semester.

Data Collection Techniques and Instruments

Data collection was carried out through the administration of a structured questionnaire to teachers at the Escuela Preparatoria Diurna of Universidad Juárez del Estado de Durango. The questionnaire was validated by experts in the field, who recommended its application. The key dimensions measured to assess the level of AI knowledge among participating teachers were: Sociodemographic information, Level of Understanding, Degree of Familiarity, Interaction Capacity, Ethical Implications, and Overall Knowledge Level—each with its respective indicators and items.

Following the administration of the questionnaires to the sample under study, it was possible to perform the necessary measurements and comparisons for this research, the results of which are discussed below:

Table 1 shows that 58.3% of the teaching staff demonstrated an intermediate level of knowledge about AI, while 20.8% reported an advanced level, and the remaining 20.8% exhibited a limited understanding. The majority, with intermediate knowledge, were familiar with general concepts and some educational applications of AI, such as classroom management tools or adaptive learning platforms. However, these teachers lacked deeper technical competencies. This aligns with findings from previous studies that highlight partial AI literacy due to limited specialized training in the educational field (Ding et al., 2024; Zhang & Zhang, 2024).

On the other hand, the group with advanced knowledge—20.8% of the total—had likely participated in technological training programs or had experience in related fields. Their background allowed them to effectively integrate AI into pedagogical environments, promoting innovative practices focused on improving learning outcomes (Tanvir et al., 2024). In contrast, the remaining 20.8% faced significant challenges in adopting these technologies, mainly due to a lack of training. This correlates with findings from a referenced study, which noted limited institutional support and negative perceptions toward AI (Sadykova & Kayumova, 2024).

Finally, factors such as age, academic background, and professional trajectory significantly influenced the level of AI proficiency.

In Table 2, familiarity with AI was predominantly low (50%), while 44.4% reported a moderate level and only 5.6% reached a high level. This knowledge gap limited not only the pedagogical appropriation of AI but also its perception as a strategic educational tool. In this regard, previous studies have confirmed that familiarity with AI is influenced by academic training, professional experience, and institutional context (Chandio et al., 2024; Sahari, 2024). Therefore, this situation highlights the urgency of implementing comprehensive and ethically informed training strategies that not only strengthen technical proficiency but also promote educators’ confidence and critical ownership to achieve optimal familiarity (Alashwal, 2024; Chandio et al., 2024).

Table 3 shows that 38.9% of the teaching staff reported a low ability to interact with AI, 52.8% reported a moderate level, and only 8.3% demonstrated a high capacity. This reveals a critical gap in preparedness for effective engagement with emerging technologies. Such disparity limits not only the functional use of AI but also hinders its transformative pedagogical integration. In this regard, Wongso et al. (2024) found that developing the ability to interact with AI systems requires a human-centered design approach. This strategy should be prioritized over traditional models, as it emphasizes usability and user experience in the development of intelligent systems. This is especially relevant given that continuous interaction models have proven to be more effective, enabling adaptive and fluid participation, thereby enhancing teacher understanding and engagement (Wintersberger et al., 2022). These findings also underscore the need to empower educators with these technologies and to promote inclusive and meaningful interaction, aligned with ethical, safety, and accessibility considerations. Neglecting these aspects may risk reinforcing existing technological inequalities (Monaro et al., 2022).

Table 4.
Ethical Implications of AI

Table 4 indicates that 38.9% of the teaching staff did not identify any ethical implications associated with the use of AI, while 45.8% acknowledged ethical concerns only occasionally. Although just 15.3% reported consistently encountering ethical dilemmas when using AI, this finding is concerning. It highlights the urgent need to establish (if not already in place) and enforce ethical frameworks for the educational use of AI, aiming to prevent opaque practices or automated decisions lacking transparency and objectivity (Simbeck, 2019). Furthermore, such frameworks should ensure privacy, informed consent, and algorithmic fairness in AI-driven decision-making (Yoğurtçu, 2024). Promoting critical faculty engagement in building a digital ethical culture is not only an urgent necessity but also a key step toward a responsible and meaningful integration of AI into educational settings (Okorie et al., 2024; Al-Adwan, 2025).

The results presented in Table 6 show that 50% of the teaching staff demonstrated a moderate level of knowledge regarding artificial intelligence (AI), while 29.2% exhibited a low level and only 20.8% reached a high level of proficiency. This distribution highlights a fragmented technological literacy that, although showing initial progress, also reveals structural gaps in the deep understanding of AI.

Basri (2024), in his study, found that the prevalence of partial or incomplete knowledge of AI represents a significant barrier to fostering sustained pedagogical transformation among educators. Conversely, those with advanced knowledge not only represent a potential strength but also a strategic human capital capable of leading educational innovation processes and peer-to-peer support (Epstein et al., 2018). Nevertheless, this is insufficient in the face of the pressing need for differentiated and progressive training models adapted to diverse professional trajectories and experience levels, where the quality of informational sources, equitable access, and ethical training play an essential role in developing critical competencies related to AI (Basri, 2024).

The results of this study indicate that the participating faculty's knowledge of artificial intelligence (AI) is predominantly at a moderate level, both in conceptual terms and in familiarity with specific tools. This finding should not be interpreted as an indicator of sufficiency, but rather as evidence of an ongoing appropriation process, conditioned by institutional, formative, technological, and human factors that globally limit the use of these technologies.

The gap observed between conceptual understanding and the effective application of AI in the classroom shows that basic knowledge alone does not ensure meaningful pedagogical integration. Likewise, the low percentage of teachers who actually use these technologies in their daily practice raises important questions about the material, symbolic, and formative conditions that mediate the adoption of AI in educational contexts. In turn, limited participation in AI-related training programs reflects structural dynamics that go beyond individual faculty responsibility.

On the other hand, although the data enabled a characterization of the current level of AI literacy, the study did not deeply address sociocultural, generational, or disciplinary variables that could influence such knowledge. Therefore, future research could expand the scope through mixed methods or longitudinal studies exploring teachers’ training trajectories, their relationship with institutional discourses on innovation, and the ethical frameworks that may or may not guide the use of intelligent technologies in education.

Finally, the strength of these findings lies not only in the quantitative diagnosis, but in the opportunity to critically examine the role that AI occupies in school environments. These results underscore, in addition to the importance of developing technical competencies for AI use and management, the need to reflect on what kind of teaching subject we seek to shape in light of AI advancement, and what role pedagogy should play in a system increasingly prone to autonomously automating educational decisions.

Conflicts of Interest
The authors declare that they have no conflicts of interest.

Author Contributions
Nolazco Piz, K. A.: Conceptualization, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review and editing.
Rodríguez Castrellón, C.: Conceptualization, Formal analysis, Investigation, Methodology, Validation, Writing – original draft, Writing – review and editing.
Casimiro Urcos, C. N.: Conceptualization, Formal analysis, Investigation, Methodology, Supervision, Validation, Writing – original draft, Writing – review and editing.

AI Use Disclosure
AI was used exclusively for grammatical revision (Spanish and English), as well as for improving the writing and style of the manuscript.