Abstract:

As part of the ACT-AI Academy (European Academy for Teachers’ Skills and Competences Development for an effective AI integration in education) EU-funded project, this study critically investigates how Generative Artificial Intelligence (AI) technologies are shaping K-12 education, specifically focusing on their role in supporting educators with learning environment design and content creation. The ACT-AI project aims to establish the ACT-AI Academy as a European hub for AI in digital education, create an AI Teacher Education Framework, develop modular AI training for pre- and in-service teachers, launch a digital learning platform for training, collaboration, and resources and boost teachers’ AI competences and awareness of AI biases through practice-based learning. Fourteen partners and three associated partners from 8 European countries that include Higher Education institutions, NGOs, SMEs, national and regional CPD providers and municipalities, participate in the project. The primary aim is to evaluate the extent to which current GenAI developments align with the pedagogical, ethical, and safeguarding requirements of primary and secondary education. The research objectives focused on synthesizing peer-reviewed evidence, mapping the functional capacities of AI tools against pedagogical needs, and evaluating the efficacy of current policy and professional development frameworks. Methodologically, the study employed a reflexive thematic analysis (Braun & Clarke, 2006) across two distinct datasets: a systematic review of peer-reviewed literature published between 2022 and 2025, and a comprehensive "Toolkit" dataset comprising AI tools, policy documents, and PD resources. The results reveal a complex landscape defined by five overarching themes. While recent empirical studies confirm that GenAI significantly enhances teacher efficiency in lesson prototyping and multimedia differentiation (Cheah et al., 2025; Lee et al., 2025), the analysis uncovers profound structural limitations. We identify a critical "governance gap" where policy frameworks remain aspirational rather than enforceable, often failing to address data privacy issues inherent in commercial tools hosted outside the EU (Giannakos et al., 2025). Furthermore, the study highlights equity concerns, noting that many tools rely on English-dominant training data and subscription models, as well as other non-inclusive practices i.e., absence of UDL principles, that risk widening the digital divide. The discussion argues that the transformative potential of GenAI is currently compromised by superficial professional development approaches that prioritize product familiarity over critical "AI literacy". We conclude that sustainable integration requires a holistic ecosystem approach—shifting from isolated experimentation to rigorous, evidence-based governance and deep pedagogical training (Celik et al., 2022). Without these systemic interventions, GenAI adoption risks exacerbating educational inequalities rather than resolving them.

Keywords:

Generative AI, K–12 Education, Content creation, Teacher Professional Development.