There is a persistent asymmetry in educational technology investment. School systems and universities invest heavily in AI tools—adaptive learning platforms, automated grading systems, intelligent tutoring software, learning analytics dashboards. They invest comparatively little in training the teachers who must use these tools. The result is predictable: expensive technology sits underutilized, or is used in ways that add administrative burden without improving pedagogy, or is adopted enthusiastically by a minority of tech-savvy faculty while the majority continues teaching as before.
The teacher professional development (TPD) literature has documented this pattern for decades across successive waves of educational technology: overhead projectors, personal computers, interactive whiteboards, learning management systems, and now AI. Each wave arrived with promises of transformation and departed leaving professional development as the unaddressed bottleneck. AI is the latest technology to encounter this pattern—but the stakes are higher because AI tools require not just technical skill but pedagogical judgment about when, how, and whether to use algorithmic recommendations.
The Bibliometric Landscape
Zhang (2025) provides a bibliometric study analyzing 217 Chinese and 500 English literature sources on AI-driven teacher professional development. AI is emerging as a transformative force in teachers' professional development within contemporary educational systems. The analysis maps the evolution of research themes and identifies convergences and divergences between Chinese and international scholarship.
The bibliometric analysis reveals that research on AI-TPD has grown exponentially since 2020, with the pandemic serving as a catalyst. But the research is unevenly distributed: most studies examine teacher attitudes toward AI or pilot programs in well-resourced contexts. Rigorous evaluation of whether AI-integrated TPD actually improves teaching practice—and student learning outcomes—remains scarce.
Rethinking TPD for the Digital Era
Anisah, Masripah, and Juanda Putri (2025) conduct a narrative review examining TPD in the digital age, particularly following the acceleration of digitalization during the COVID-19 pandemic. The study analyzes best practices, challenges, and strategies for effective TPD by reviewing literature from Scopus, Web of Science, and ERIC databases.
The review identifies a fundamental mismatch: traditional TPD models (one-day workshops, summer institutes, certification courses) were designed for stable pedagogical knowledge that changes slowly. AI tools change rapidly—new capabilities emerge quarterly, interfaces redesign frequently, and the pedagogical implications of each update require fresh evaluation. Teachers trained on a specific AI tool in September may find that tool substantially different by January.
Effective TPD for the AI era, the review suggests, requires:
- Continuous, embedded learning: Training integrated into daily practice rather than isolated in workshops
- Peer learning communities: Teachers learning from each other's AI integration experiences
- Just-in-time support: Access to help when facing specific AI-related challenges, not months later in a scheduled training session
- Critical AI literacy: Understanding not just how to use AI tools but when to override them, when to distrust them, and when to choose human judgment over algorithmic recommendation
Basic Education: The Foundational Challenge
Schwarzbach, Vicari, and Ribeiro (2025) review publications on AI in basic education (K-12), noting a global increase in interest with China and the USA leading contributions. Prominent research areas include teacher perceptions of AI integration and the practical challenges of implementation.
The basic education context presents challenges that differ from higher education. K-12 teachers typically have less autonomy over curriculum and technology choices, less access to professional development funding, and larger class sizes that make individualized AI tool adoption more difficult. The "teacher perception" focus in the literature reflects a reality: much AI implementation in K-12 is decided by administrators and deployed to teachers, rather than chosen by teachers based on their pedagogical judgment.
Engineering Faculty: Industry 4.0 Pressure
Fernández Cerero et al. (2026) analyze AI adoption and educational uses among engineering faculty, where the pressure to integrate AI comes not only from educational reform but from industry expectations. AI has established itself as a key technology in the context of Industry 4.0, with direct implications for university education.
Engineering faculty face a distinctive challenge: they must prepare students for an AI-augmented professional practice while their own pedagogical training may not include AI. The study examines the degree of adoption and the main educational uses of AI-based tools in higher education, revealing that engineering faculty who use AI in their research are more likely to integrate it into their teaching—but that research use and pedagogical use require different competencies.
The Policy Gap: Pakistan vs. Australia
Ishaq, Iqbal, and Akhar (2025) provide a comparative study of educational leaders' experiences with AI integration in Pakistan and Australia. Using semi-structured interviews with 16 educational leaders, the study captures how institutional contexts shape AI adoption.
The comparison reveals that the policy infrastructure for AI in education differs dramatically between the two countries. Australia has national AI ethics frameworks, funded digital literacy programs, and professional standards that reference technology integration. Pakistan has limited formal policy guidance, minimal funding for AI-specific TPD, and a digital divide that makes AI integration irrelevant for many schools.
The policy gap finding is generalizable beyond these two countries: effective AI integration in education requires a policy ecosystem that includes infrastructure investment, professional development funding, quality standards for AI tools, and ethical guidelines for AI use in classrooms. Where this ecosystem is absent, individual teacher initiative cannot compensate for systemic shortcomings.
Claims and Evidence
<| Claim | Evidence | Verdict |
|---|---|---|
| Teacher professional development is the primary bottleneck for AI in education | All five papers identify TPD as a critical barrier; technology availability alone is insufficient | ✅ Supported |
| Traditional TPD models are adequate for AI integration | Anisah et al. (2025): one-day workshops and certification courses are insufficient for rapidly changing tools | ❌ Refuted |
| Teacher attitudes toward AI are the main adoption barrier | Zhang (2025), Fernández Cerero et al. (2026): attitudes matter but structural factors (time, training, support) matter more | ⚠️ Uncertain (attitudes are necessary but not sufficient) |
| AI integration policy is consistent across Global North and South | Ishaq et al. (2025): dramatic policy infrastructure gaps between Australia and Pakistan | ❌ Refuted |
Open Questions
Implications
The research converges on a practical conclusion: the return on investment in educational AI depends more on teacher professional development than on the technology itself. A well-trained teacher with a basic AI tool will produce better outcomes than an untrained teacher with a sophisticated one. Until education systems invest in TPD at a level commensurate with their investment in technology, AI's educational impact will remain far below its potential.