Research Article
Teaching the teachers: A systematic review of genAI-specific technological pedagogical knowledge (TPK) in teacher education
Synthesis: Ó Ceallaigh and Murphy (2026) systematically review the GenAI-specific Technological Pedagogical Knowledge (TPK) required by teacher educators (TEs) who prepare post-primary teachers — an underexplored area — following PRISMA guidelines. The final search (April 2025) covered ERIC and EBSCO for English-language peer-reviewed empirical studies from 2022–2025 (Scopus and Web of Science were excluded due to record overlap), using a modified PIO framework (Population, Intervention, Outcome) in place of PICO because comparison groups are uncommon in education research. Seventeen studies met the inclusion criteria, spanning qualitative, quantitative, and mixed-methods designs — e.g., Celik's Intelligent-TPACK scale development in Finland, Moorhouse & Kohnke's 604-instructor survey, Nyaaba & Zhai's 307 teacher educators in Ghana, and Warr & Heath's audit of LLM feedback bias. The synthesis finds that GenAI-specific TPK extends beyond digital competence to encompass pedagogical reasoning, ethical awareness, and AI-augmented learning design, with persistent challenges around ethical decision-making, critical AI literacy, data privacy, algorithmic bias, authorship, and accountability — issues especially salient in teacher preparation within higher education institutions (HEIs). The authors argue HEIs must foster coherent strategies, supportive policies, and sustained professional learning to build TEs' GenAI-TPK.
Key Findings
Scope and design. The review answers three research questions on the TPK pre-/in-service post-primary teachers need, the TPK TEs require to prepare them, and how TEs can be institutionally and professionally supported. Screening used a standardized coding framework informed by Miles & Huberman's two-phase model — vertical (within-case) analysis of each of the 17 studies followed by horizontal (cross-case) thematic synthesis; exclusion criteria removed non-empirical, non-peer-reviewed, and primary-level-focused work.
TPK extends beyond digital competence. Framed within TPACK (rooted in Shulman's PCK, extended by Mishra & Koehler) and augmented by Intelligent-TPACK (adding an ethics domain), Contextual Knowledge (XK), and the Enabling AI framework, GenAI-TPK for TEs is conceptualized as context-dependent and dynamic, requiring pedagogical judgment, critical evaluation, and ethical reasoning embedded within — rather than separate from — technology integration.
Included-studies evidence. The 17 studies show varied educator readiness: e.g., Barrett & Pack (68 educators, 158 students) found teachers lacked guidance for ethically integrating ChatGPT into writing education pedagogy; Dilling & Herrmann documented emergent "instrumental genesis" as preservice math education teachers repurposed ChatGPT for proofs; Kong et al. showed an AI-literacy course improved 128 secondary students' ethical understanding; Guan et al. interviewed 24 pre-service K-12 teachers on reconceptualizing teacher identity for AI collaboration; Lee et al. surveyed 30 Australian university educators facing inconsistent institutional support; Navío-Inglés et al. had 154 Spanish teacher education rate AI-vs human-written texts, with many failing to identify AI authorship; Wang et al. found most U.S. universities adopt "open-but-cautious" GenAI stances emphasizing academic integrity; and Warr & Heath's audit found LLM feedback reproduced systemic inequities (e.g., lower scores for marginalized identities).
Institutional responsibility and HEI role. The review stresses that HEIs commonly offer pedagogical courses and professional development — accelerated since the Covid-19 pandemic — yet there is little evidence of how teacher-education programs specifically prepare future educators for GenAI. Fostering coherent institutional strategies, supportive policies, and sustained, subject-relevant professional learning is framed as central to enabling TEs' GenAI-TPK.
Ethical challenges are central. Themes of bias/fairness, accountability/transparency, privacy/data ethics, student Over-Reliance, and the need for ethical literacy recurred across studies, calling for a shift from reactive policy toward critical, reflective, equity-centered GenAI education, including explainable AI (XAI) and prompt engineering training.
Limitations. The small number of included studies reflects the field's novelty and strict PRISMA criteria; restricting the search to ERIC and EBSCO may have excluded relevant work from Scopus, Web of Science, or Google Scholar; publication bias, limited heterogeneity, time-lag bias, and the interpretive nature of thematic synthesis (no meta-analysis or effect-size aggregation was possible) constrain generalizability.
Implication. Teacher-education programs must explicitly build teacher educators' GenAI-TPK, with HEIs enabling this through coherent strategies, supportive policies, and sustained professional learning — connecting to teacher education, TPACK, ethics, curriculum design, and educational development.
What this means for practice
- Teacher educators. Make GenAI-specific TPK an explicit, assessed element of initial teacher education — pedagogical reasoning, ethical awareness, and AI-augmented learning design — rather than leaving it to incidental digital competence.
- Teacher educators. Teach ethical decision-making as part of technology integration: bias and fairness, accountability and transparency, privacy and data ethics, and student over-reliance recurred across the included studies.
- Faculty developers. Design sustained, subject-relevant professional learning instead of one-off courses; the review found little evidence of how teacher-education programs specifically prepare future educators for GenAI and reports inconsistent institutional support.
- Administrators. Establish coherent institutional strategy and supportive policy before expecting staff to integrate GenAI, since the reviewed evidence points to fragmented guidance rather than systemic preparation.
Limitations
- Only 17 studies met the inclusion criteria (English-language peer-reviewed empirical work, 2022–2025), a small corpus the authors attribute to the field's novelty and their strict application of PRISMA criteria.
- The search was confined to ERIC and EBSCO; Scopus and Web of Science were excluded because of record overlap, so relevant interdisciplinary or international work may have been missed.
- With mostly qualitative designs, no meta-analysis or quantitative aggregation was feasible — no effect sizes or trend estimates can be drawn from the synthesis.
- Publication bias is likely (positive findings are more publishable than null results) and heterogeneity among the included studies is limited, constraining generalizability.
Citation
Ó Ceallaigh, T. J., & Murphy, S. (2026). Teaching the teachers: A systematic review of genAI-specific technological pedagogical knowledge (TPK) in teacher education. Computers and Education Open, 100367. https://doi.org/10.1016/j.caeo.2026.100367