This paper examines the development of Artificial Intelligence (AI) literacy from a teacher-centered perspective in Italy. Despite the growing importance of AI in education, its integration into school curricula remains a major challenge. Key barriers include complex terminology, limited opportunities for hands-on professional development, and the absence of clear national curricular guidelines. The study identifies the main obstacles to developing AI literacy among in-service teachers, explores the tools and strategies currently adopted in professional training, and analyzes the objectives and content of AI-related courses ofered through Scuola Futura, the national teacher training platform established under Italy's National Recovery and Resilience Plan (PNRR), as well as university programs for pre-service teachers. Findings show that most training initiatives focus on introductory and general digital skills rather than advanced or discipline-specific applications. Although recent national measures such as Decreto Ministeriale (D.M.) 66/2023 represent a positive step, sustained eforts are still required to bridge the AI knowledge gap and promote meaningful, pedagogically grounded integration of AI in teaching practice.
Artificial Intelligence (AI) is taking on growing importance in educational agendas, highlighting the
need for stronger teacher preparation. However, there is no universal definition, as the term reflects
diverse interpretations shaped by the pervasive role of AI in society. It generally means having the
skills to critically evaluate, communicate about, and use AI tools ethically in various contexts. European
policies identify AI literacy as essential for everyone [
© 2025 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
corresponding to the "acquire level" in the proposed framework [
The paper is organized into three main sections. The first provides an overview of the current state of research on AI literacy in education, while the second and third present an analysis of the courses and tools employed by in-service and pre-service teachers to foster AI competencies within the Italian educational context.
In recent years, several European initiatives have been launched to promote AI literacy in education.
Among them, the AI4T project (Erasmus+ K3, 2021) stands out as a major collaborative efort in Europe
that aims to improve the abilities of teachers related to AI. The project was structured around three key
domains: Teaching for AI (general competencies aligned with the DigComp 2.2 framework), Teaching
with AI (teacher-specific digital and pedagogical skills) [
Despite the rapid expansion of both online [
During the COVID-19 pandemic, Massive Open Online Courses (MOOCs) were widely adopted,
acting as a catalyst for digital transformation and online professional learning. Although this experience
familiarized most educators with digital platforms, the integration of AI into teaching practice has
progressed more slowly, as many still lack confidence and hands-on experience with AI-based tools
[
Acceptance or resistance to AI in educational technology (AI EdTech) is strongly influenced by teachers’ AI literacy [17], which is essential to mitigate biases and misconceptions [18]. In fact, some educators integrate AI into pedagogy through exploration, critical evaluation, and collaborative work [19]. For a responsible adoption, it is essential to know how to evaluate tools in terms of reliability, performance, precision, and ethical implications. Educators with strong AI literacy should collaborate with AI specialists in co-designing curricula, teaching strategies, and learning activities that meaningfully integrate AI into classroom practice [20]. An ideal opportunity to foster this collaboration is during PATH, a summer school program promoted by INDIRE, the Italian National Institute for Educational Research and Innovation, which supports experimentation and teacher professional development. In addition, research [16] shows that educators are more likely to explore and adopt AI tools when they feel autonomous in their choices and confident in their ability to use them. This sense of competence and control fosters greater motivation and openness to innovation.
AI ofers opportunities to address teaching challenges and engage students in real-world problem
solving [21], but this requires developing both pedagogical and technical skills. Institutional support
plays a crucial role in meeting teachers’ psychological needs for autonomy and competence, thereby
fostering AI literacy. Equipping educators with foundational AI skills promotes lifelong professional
growth and enables them to contribute to the future of society [
For educators, robust information is crucial to efectively integrate AI into teaching and assessment. This involves guiding students in managing information through AI, developing digital resources, and applying diverse systems in instructional design, while also fostering ethical and privacy-conscious use of AI in communication and collaboration. Moreover, integration can be supported by accessible plug-and-play tools designed to introduce AI concepts in education, such as PictoBlox and mBlock, which are preferably free or provide open-access versions, without registration to ensure accessibility and avoid account-related issues.
Our analysis of in-service teacher training courses presented in Scuola Futura [22] and DIDACTA workshops 2025 [23] suggests the following recommended tools: • Hands-on lab activities: Teachable Machine and Quick, Draw! (by Google), as well as AI-based applications on Code.org such as AI for Oceans. • General-purpose use: ChatGPT, Copilot, Gemini AI, and NotebookLM, available in Google
Workspace for Education suite. • Image generation: DALL·E, Gemini AI, Padlet, and Leonardo.ai. • Music creation: Suno. • AI-enhanced presentations: Canva AI and Gamma.
• Lesson planning and advanced learning design: Linda and AIforL (by FEM).
A critical factor in the adoption of AI in education is not only the understanding of the technology by educators, but also the degree of trust and perceived transparency that AI systems inspire. Indeed, teachers are often reluctant to use AI-based tools because they do not trust the information provided [24]. This is largely due to the perception of AI as a “black box”, making it dificult for them to understand how and why algorithms generate certain outputs or make specific decisions. For this reason, AI literacy is essential to explain in a clear and accessible way how AI systems operate. In this context, Explainable Artificial Intelligence (XAI) [ 25] focuses on making AI systems more transparent and interpretable. To promote ethical and flexible AI use in education, schools and teachers should adopt a vendor-neutral, “agnostic approach” [26], choosing models and tools that align with pedagogical goals and uphold transparency and interpretability.
Trust in AI is a dynamic phenomenon [27], influenced by multiple factors such as context, frequency of use, prior experience with the system, and the credibility of the institution promoting it (e.g., Platone AI platform sponsored by INDIRE [28]). Indeed, teachers are more likely to trust and integrate AI tools when they align with their pedagogical approaches, contribute to workload reduction, and are developed or supported by credible institutional providers. In general, teachers approach automated tools with considerable skepticism, but ensuring the pedagogical quality can be a key factor in building greater confidence [29].
To gain a comprehensive overview, this study conducted a structured analysis of the courses available on Scuola Futura. The platform provides free professional development for teachers, school staf, and principals, accessible through the national authentication systems Sistema Pubblico di Identità Digitale (SPID) or Italian electronic ID system (CIE), and delivered in in-person, hybrid, or predominantly online formats.
Data collection took place between late July and the first week of August 2025. The search was carried out directly through the Scuola Futura public interface using the keywords "AI" and "intelligenza artificiale" within the filtering category "Transizione digitale". No relevant results were retrieved when searching for "generative AI" or "GenAI". The initial dataset consisted of 100 active courses available nationwide. Each course entry included information such as title, short description, training modality, and geographic location of the hosting institution. After removing duplicates and inactive listings, a manual screening was performed to assess thematic relevance. Courses that did not explicitly mention AI, or that focused solely on generic digital literacy, were excluded. This process yielded a final dataset of 63 unique courses distributed across all Italian regions, covering multiple education levels, from primary to secondary teachers’ training. The results highlight a strong and growing interest in AI-supported teaching, with a particular focus on themes such as inclusion, creativity, and emerging technologies. The data analysis followed a manual content-coding procedure applied to course titles and descriptions. The process involved the following steps: 1. Identify recurring keywords and concepts emerging from course titles and descriptions; 2. Cluster similar items into broader thematic categories; 3. Analyze and quantify the occurrence of each theme, and generate visual representations using
Python Matplotlib.
This approach allowed the identification of four main macro-categories as illustrated in Figure 1. • General pedagogy (e.g., “insegnare con AI ”, “curricolo digitale”): predominantly linked to terms such as “strumenti”, “strategie”, “MOOC” and “sfide ”. • Machine Learning and technical AI: focused on developing a technical understanding of algorithms and deep models. • Inclusion (e.g., BES) and accessibility: central in courses on personalized learning pathways. • Storytelling and digital creativity: aimed at fostering student engagement through narrative and creative approaches.
Geographical distribution and course typologies were then examined to highlight national trends and potential imbalances between regions. This analysis revealed a concentration of initiatives in schools located in both Northern and Southern Italy, suggesting that AI literacy programs are increasingly perceived as a nationwide priority. All data were drawn from publicly available sources; no personal or sensitive information was collected. However, it is important to note that the analysis does not include locally organized training initiatives that are not publicly listed on Scuola Futura, representing a possible limitation of the presented dataset. Among the 100 courses initially considered, only a few focus on emerging topics such as educational robotics, metaverse, VR/AR, blockchain, NFTs, and digital security and privacy. The predominance of general courses on digital education and AI indicates that the current training provision is primarily oriented toward fostering a foundational level of AI literacy, focusing more on practical use than on a deep understanding of AI. However, the growing presence of courses on machine learning (as illustrated in Figure 2), GenAI, and tools indicates increasing interest in pedagogical approaches augmented by AI technologies. This trend underscores the need to invest in more advanced, technically oriented training, particularly for STEM educators, while progressively extending AI literacy to non-STEM subjects. National policies could therefore promote interdisciplinary training models, certification pathways in AI literacy, and the creation of regional AI labs where teachers can collaboratively experiment with AI tools in authentic learning contexts. Moreover, Figure 3 highlights the importance of moving beyond isolated, project-based initiatives by embedding AI literacy systematically within national education policies. Whereas current programs often remain fragmented or limited to temporary projects (such as those funded under the PNRR), a coherent policy framework is required to ensure the long-term integration of AI literacy into teacher education at all levels. Such an approach would foster not only advanced, technically oriented training but also the gradual inclusion of AI-related pedagogies across the humanities, social sciences, and arts, enabling teachers to address the ethical, societal, and creative dimensions of AI.
To assess the integration of AI literacy into future teacher training, we examined the courses ofered by accredited Italian universities to deliver teacher qualification programs, as listed in the oficial register of the Ministry of University and Research [31]. Specifically, we focused on 25 universities that ofer the qualification for the A041 teaching class, which refers to Computer Science. However, to the best of our knowledge, and based on information available on public university webpages, the analysis revealed that only 2 out of 25 universities ofer courses explicitly dedicated to AI: • “Laboratorio di intelligenza artificiale e cybersecurity nei contesti educativi ” at the University of
Venice.
• “Pillole di reti neurali” at the University of Udine.
These findings highlight the need for more balanced national planning and systematic consolidation of AI competencies throughout the school system, targeting both in-service and pre-service teachers [32], and extending AI education in and beyond computer science subjects.
This study examined AI literacy in the Italian educational system through an analysis of training opportunities for in-service and pre-service teachers. Despite emerging initiatives and increasing awareness, AI education remains largely introductory, with limited focus on advanced technical skills or cross-disciplinary applications.
In addition, current AI-based tools are often limited to single modalities, such as text input, which constrains personalization and creativity. Expanding toward multimodal systems that integrate images, sound, video, and text could enable more adaptive and engaging learning experiences, as demonstrated in studies with pre-service teachers using generative AI for project-based exploration [21].
A significant challenge remains the absence of a shared framework for assessing AI literacy [ 33]. The paper argues for clear policy guidelines aligned with curricula and systematic integration of AI literacy to develop critical, creative, and ethical AI competencies among educators and, consequently, their learners.
This paper was partially supported by University of Udine project on “Piano Stretegico Dipartimentale on Artificial Intelligence” (PSD-AI) (2022-25) project at the University of Udine.
During the preparation of this work, the authors used GPT-5 in order to grammar and spelling check. After using these tools, the authors reviewed and edited the content as needed and assumed full responsibility for the publication’s content. stumbling stones, Interact. Des. Archit. 47 (2021) 48–68. URL: https://hdl.handle.net/11250/3018456. doi:10.55612/s-5002-047-003. [16] T. Nazaretsky, M. Ariely, M. Cukurova, G. Alexandron, Teachers’ trust in AI-powered educational technology and a professional development program to improve it, Br. J. Educ. Technol. 53 (2022) 914–931. URL: https://doi.org/10.1111/bjet.13232. [17] F. Sun, P. Tian, D. Sun, Y. Fan, Y. Yang, Pre-service teachers’ inclination to integrate AI into STEM education: Analysis of influencing factors, Br. J. Educ. Technol. 55 (2024) 2574–2596. doi:10.1111/bjet.13469. [18] B. Williamson, F. Macgilchrist, J. Potter, Re-examining AI, automation and datafication in education,
Learn. Media Technol. 48 (2023) 1–5. doi:10.1080/17439884.2023.2167830. [19] C. Nerantzi, S. Abegglen, M. Karatsiori, A. Martínez-Arboleda, 101 creative ideas to use AI in education: A crowdsourced collection (2023 1.0), Zenodo, 2023. doi:10.5281/zenodo.8072950. [20] S. Benevenuta, R. Treglia, AI computational and ethical awareness in italian high schools: A series of workshops, in: Proc. 1st Workshop on Education for Artificial Intelligence (EDU4AI 2024), co-located with the 23rd Int. Conf. of the Italian Association for Artificial Intelligence (AIxIA 2024), Bolzano, Italy, 2024. URL: https://ceur-ws.org/Vol-3902/14_paper.pdf. [21] B. E. Smith, A. Y. Shimizu, S. K. Burriss, M. Hundley, E. Pendergrass, Multimodal composing with generative AI: Examining preservice teachers’ processes and perspectives, Comput. Compos. 75 (2025) 102896. doi:10.1016/j.compcom.2024.102896. [22] M. dell’Istruzione e del Merito, Percorsi formativi – scuola futura (pnrr), 2025. URL: https:// scuolafutura.pubblica.istruzione.it/percorsi. [23] INDIRE, Fiera didacta italia, https://fieradidacta.indire.it/it/, 2025. Accessed: 13 October 2025. [24] Y. Feldman-Maggor, M. Cukurova, C. Kent, et al., The impact of explainable AI on teachers’ trust and acceptance of AI EdTech recommendations: The power of domain-specific explanations, Int. J.
Artif. Intell. Educ. (2025). doi:10.1007/s40593-025-00486-6. [25] European Commission: European Education and Culture Executive Agency, Explainable ai in education – fostering human oversight and shared responsibility, 2025. URL: https://data.europa. eu/doi/10.2797/6780469. [26] UNESCO, K-12 ai curricula: A mapping of government-endorsed ai curricula, UNESCO, 2022. URL: https://doi.org/10.54675/ELYF6010. [27] E. Anagnostopoulou, N. A. Grammatikos, D. Apostolou, G. Mentzas, Trustworthy AI in education: A roadmap for ethical and efective implementation, in: Proc. 13th Hellenic Conf. on Artificial Intelligence (SETN ’24), Association for Computing Machinery, New York, NY, USA, 2024. doi:10. 1145/3688671.3688781. [28] INDIRE, Platone ai: dialoga con platone attraverso l’intelligenza artificiale, 2024. URL: https: //platoneai.indire.it/. [29] E. Marengo, M. Baldoni, Non è mai troppo tardi 2.0, Technical Report, Dipartimento di Informatica,
Università degli Studi di Torino, 2020. [30] M. d. Unità di Missione per il PNRR, Didattica digitale integrata e formazione sulla transizione digitale del personale scolastico – missione 4 c1, investimento 2.1, 2024. URL: https://pnrr.istruzione.it/competenze/ didattica-digitale-integrata-e-formazione-sulla-transizione-digitale-del-personale-scolastico/. [31] Ministero dell’Università e della Ricerca (MUR), Elenco percorsi accreditati in relazione alle classi di concorso per ciascuna istituzione, 2024. URL: https://www.mur.gov.it/sites/default/files/2024-02/ ELENCO%20CLASSI%20ACCREDITATE%20PER%20CIASCUNA%20ISTITUZIONE.pdf. [32] Y. Zhang, C. Lai, M. M. Y. Gu, Becoming a teacher in the era of AI: A multiple-case study of pre-service teachers’ investment in AI-facilitated learning-to-teach practices, System 133 (2025) 103746. doi:10.1016/j.system.2025.103746. [33] R. Xiao, Y.-J. Tseng, H. Li, G. Liao, J. Stamper, K. R. Koedinger, AI literacy for all: 1st international workshop on AI literacy education for all, in: Artificial Intelligence in Education. Posters and Late Breaking Results, Workshops and Tutorials, Industry and Innovation Tracks, Practitioners, Doctoral Consortium, Blue Sky, and WideAIED (AIED 2025), Springer, Cham, 2025, pp. 255–260.
BES CIE INDIRE MOOC PNDS PNRR SPID STEM UNESCO XAI
Ministerial Decree, oficial directive issued by the Italian Ministry of Education under the PNRR framework (e.g., D.M. 65/2023,
D.M. 66/2023) Bisogni Educativi Speciali Special Educational Needs, referring to students requiring personalized educational support Carta d’Identità Elettronica Italian Electronic Identity Card used for secure online authentication in public services Digital Competence Framework for European framework developed by the Educators Joint Research Centre (JRC) to define edu
cators’ digital competencies.
Generative Artificial Intelligence Branch of AI focused on the autonomous generation of text, images, and other creative content Istituto Nazionale di Documen- Italian National Institute for Educational tazione, Innovazione e Ricerca Research and Innovation Educativa Massive Open Online Course Piano Nazionale Scuola Digitale
The dataset used in this study can be accessed via an open-access repository