Teachers still resist AI despite training: Here's the missing link

Artificial intelligence (AI) is steadily entering classrooms worldwide, promising to reshape lesson planning, assessment, and student engagement. Yet despite growing access to AI tools and training programs, a critical gap remains between availability and actual classroom use. New global research suggests that the missing link is not technology or even training, but teachers' beliefs about AI itself.

A study published in Education Sciences analyzes how professional development and psychological factors interact to shape AI use in education. Titled "Global Evidence on the Mediating Role of Teachers' AI Beliefs in Linking AI-Related Professional Development and AI Infusion in Instruction," the research offers new insights into why AI adoption in classrooms remains uneven despite rapid technological advancement.

Based on data from 34,628 teachers across 55 economies participating in the 2024 Teaching and Learning International Survey (TALIS), the study moves beyond surface-level explanations of AI adoption and identifies a deeper behavioral mechanism. While professional development directly improves both AI use and attitudes, it is teachers' beliefs that ultimately determine whether AI becomes embedded in daily teaching practice.

Professional development drives skills but not automatic adoption

The study confirms that AI-related professional development plays a foundational role in preparing teachers for digital transformation. Training programs expose educators to AI tools, pedagogical applications, and technical skills, helping them understand how these technologies can support instruction.

Statistical analysis shows a clear and significant relationship between participation in AI-focused training and teachers' beliefs about AI, as well as their actual use of AI in classrooms. Teachers who received professional development were more likely to view AI as useful and were also more likely to integrate it into instructional activities.

This aligns with broader trends in educational technology, where structured learning opportunities improve teachers' confidence and readiness to adopt new tools. AI training, in particular, helps educators move beyond abstract awareness by providing hands-on experience with applications such as lesson planning, content generation, and automated assessment.

However, the study highlights a critical limitation. Training alone does not guarantee meaningful integration. Many teachers who receive professional development still hesitate to use AI consistently in their teaching. This disconnect reflects a deeper issue: the transition from knowledge to practice depends on how teachers interpret and trust the technology.

The research notes that professional development functions as an enabling condition rather than a complete solution. It equips teachers with the skills to use AI, but whether those skills translate into sustained classroom use depends on additional psychological and contextual factors.

Teachers' beliefs shape the real adoption of AI in classrooms

Teachers' beliefs about AI act as the strongest predictor of whether the technology is actually used in instruction. Educators who perceive AI as useful, reliable, and aligned with pedagogical goals are significantly more likely to integrate it into their daily teaching practices.

The data reveals a strong and statistically significant relationship between positive AI beliefs and instructional use, with belief systems exerting a much larger influence than training alone.

These beliefs encompass multiple dimensions, including perceived usefulness, trust in AI outputs, confidence in using the technology, and alignment with teaching objectives. Teachers who see AI as a meaningful instructional aid are more likely to use it for tasks such as adapting learning materials, supporting diverse student needs, and automating administrative processes.

On the other hand, teachers who harbor doubts about AI's reliability or relevance are less likely to adopt it, even if they have received training. Concerns about accuracy, ethical implications, and classroom appropriateness continue to act as barriers to integration.

The findings reinforce the idea that technology adoption in education is not purely technical but deeply psychological. Teachers do not simply implement tools because they are available or because they have been trained to use them. Instead, they evaluate these tools through a lens of professional judgment, weighing their perceived value against potential risks.

This dynamic helps explain why, despite widespread awareness of AI technologies, actual classroom use remains limited in many contexts. Awareness does not automatically translate into acceptance, and acceptance does not guarantee sustained use.

Beliefs act as the bridge between training and classroom practice

The researchers identify teachers' beliefs as a mediating factor between professional development and AI infusion. The analysis shows that a substantial portion of the impact of training on AI use operates indirectly through changes in teachers' beliefs.

Approximately one-third of the effect of professional development on AI integration is mediated by belief systems, indicating that training influences classroom practice not only by building skills but also by reshaping how teachers perceive AI.

This mediation effect highlights a critical pathway in the adoption process. Professional development introduces teachers to AI tools and demonstrates their potential applications. These experiences then influence teachers' internal attitudes, including their confidence, trust, and perceived usefulness of AI. It is these revised beliefs that ultimately drive consistent and meaningful use in instructional settings.

Practically, this means that even well-designed training programs may fail to achieve their goals if they do not address teachers' perceptions and concerns. Technical competence alone is insufficient. Teachers must also develop a positive and informed understanding of AI's role in education.

The study places these findings within the Unified Theory of Acceptance and Use of Technology (UTAUT), which emphasizes the role of performance expectancy, effort expectancy, and facilitating conditions in shaping technology adoption. In this framework, professional development enhances both competence and perceived ease of use, while beliefs influence behavioral intentions and actual usage.

By integrating these elements, the research provides a more comprehensive explanation of how AI adoption unfolds in educational contexts. It moves beyond simple cause-and-effect models to reveal a layered process in which training, beliefs, and practice interact dynamically.

Global data reveals uneven but emerging AI integration trends

The study's global scope offers a rare cross-national perspective on AI adoption in education. Drawing on TALIS 2024 data, it captures a diverse sample of teachers from different regions, age groups, and professional backgrounds.

The dataset shows that the majority of teachers are mid-career professionals, with a significant proportion working full-time in permanent roles. This demographic profile suggests that the findings reflect mainstream teaching populations rather than early adopters or niche groups.

Across this diverse sample, consistent patterns emerge. Professional development is widely recognized as a key driver of AI readiness, but its impact varies depending on how teachers interpret and internalize their experiences. Beliefs consistently emerge as a central factor influencing whether AI becomes part of routine teaching practice.

The findings also point to broader systemic challenges. Educational institutions often focus on providing access to technology and training without addressing the underlying attitudes that shape adoption. As a result, investments in AI infrastructure may not yield the expected outcomes if teachers remain uncertain or skeptical about its value.

Policy and practice implications for AI in education

The study suggests that successful AI adoption requires a more holistic approach that combines technical training with efforts to shape teachers' perceptions and confidence.

Professional development programs must go beyond skill-building to include opportunities for reflection, discussion, and critical engagement with AI technologies. Teachers need to understand not only how to use AI but also why it matters and how it aligns with their instructional goals.

This includes addressing common concerns about reliability, ethics, and classroom impact, as well as providing practical examples of successful integration. Collaborative learning environments, peer support, and ongoing feedback can help reinforce positive beliefs and reduce resistance.

The research also highlights the importance of supportive institutional conditions, including access to resources, time for experimentation, and leadership support. These factors create an environment in which teachers feel empowered to explore and adopt new technologies.

A turning point for AI in teaching and learning

Education systems face a critical moment in determining how these technologies will be integrated into teaching and learning. The study makes clear that the future of AI in education will not be shaped by technology alone but by the people who use it.

Teachers remain the key agents of change. Their beliefs, attitudes, and professional judgments will determine whether AI becomes a transformative force or remains an underutilized tool.

The findings suggest that bridging the gap between potential and practice requires a shift in focus. Instead of viewing AI adoption as a purely technical challenge, education systems must recognize it as a human-centered process that depends on trust, confidence, and alignment with pedagogical values.

In this context, professional development is not just about teaching skills but about shaping mindsets. By addressing both dimensions, education systems can create the conditions for AI to move from experimentation to sustained, meaningful integration in classrooms worldwide.