Teachers Interacting with Generative Artificial Intelligence: A Dual Responsibility Carmine Gravino†,1, Alessandro Iannella*,†,2,3 , Mirko Marras†,3, Silvio Marcello Pagliara†,3, Fabio Palomba†,1 1Università degli Studi di Salerno 2Università degli Studi di Macerata 3Università degli Studi di Cagliari Abstract This paper explores the integration of Generative Artificial Intelligence (GAI) in education, focusing on the pivotal role of teachers. It introduces current policies from educational organizations and governmental institutions, and briefly reports on research evidence on the topic, emphasizing the need for targeted training programs. These programs are essential for equipping educators with proactive strategies to harness this emerging technology, not only to enhance professional practice but also to effectively guide students in developing AI literacy, utilizing both hidden and explicit curricula. Keywords education, GAI systems, teacher training. 1. Introduction policy documents [5] and is currently developing a new AI competency framework for teachers and The educational landscape is currently school students [6]. Within the academic context, undergoing a significant transformation, driven by there are significant investments concerning the rapid advancements in technology. The advent of drafting of regulations and directives, the launch of ChatGPT and other Generative Artificial Intelligence experimental studies [7], and the creation of themed (GAI) systems marked the beginning of a period laboratories. characterized by increasing interest in Artificial The educational sector is generally characterized Intelligence (AI). by its slow pace. It takes time to observe the outcomes Educational organizations and governmental of implemented strategies. However, it is currently a institutions, as well as academic communities, have critical area for reflection and the deployment of AI actively responded to this emerging technology, and GAI technologies, facing the challenge of engaging continuing a trajectory that began years earlier with with a subject that evolves even as it is being tutoring systems [1], personalized learning [2], and discussed [8]. the collection and analysis of learning data [3]. This paper examines proposals for teacher European projects commissioned and coordinated by training, considering input from educational the European Union are producing resources, organizations and governmental institutions at global, guidelines, and reports [4]. UNESCO issued several Ital-IA 2024: 4th National Conference on Artificial Intelligence, . organized by CINI, May 29-30, 2024, Naples, Italy 0000-0002-4394-9035 (C. Gravino); 0000-0003-1533-1884 ∗ Corresponding author. (A. Iannella); 0000-0003-1989-6057 (M. Marras); 0000-0002- † These authors contributed equally. 0175-5160 (S. M. Pagliara); 0000-0001-9337-5116 (F. Palomba). gravino@unisa.it (C. Gravino); © 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). a.iannella@unimc.it (A. Iannella); mirko.marras@unica.it (M. Marras); silviom.pagliara@unica.it (S. M. Pagliara); fpalomba@unisa.it (F. Palomba). CEUR ceur-ws.org Workshop ISSN 1613-0073 Proceedings European, and Italian levels, as well as evidence from evaluate the quality and relevance of the system's scientific research. The aim is to stimulate reflection output. From this perspective, GAI systems can be on the dual importance of such training, which is viewed as process partners rather than content relevant not only for enhancing professional practice partners, assisting teachers in instructional practices but also for guiding students in developing AI literacy. and students in their own learning. In the latter case, it might be more appropriate that GAI systems 2. Generative Artificial Intelligence in support methodological and metacognitive aspects of learning rather than the exploration of a specific Education disciplinary content. GAI is a specific type of AI within the area of GAI systems serve as evocative objects of Machine Learning (ML). It aims at creating new individual knowledge too, as they necessitate ongoing content such as texts, images, videos, music, and 3D reflection during the formulation of prompts and models, based on the data learned from during a analysis of outputs, encouraging a deep consideration training phase. GAI uses several key technologies, of what is known and what remains unknown [9]. including Generative Adversarial Networks (GANs), Recurrent Neural Networks (RNNs), and 3. Education in the Italian AI Strategy transformers like the Generative Pre-trained Transformer (GPT) models. for 2024-2026 The interaction process with GAI systems, Italy is aligning with the European approach to AI especially with Large Language Models (LLMs), is through a new AI Strategy. In April 2024, the mostly conversational. Dialogue exemplifies a shared Department for Digital Transformation and the domain of knowledge where the human partner Agency for Digital Italy released an executive progressively “reads” the algorithmic knowledge of summary regarding the period 2024-2026 [17]. The the machine through his/her own internal knowledge previous strategy covered the period 2022-2024 [18]. [9]. The summary encompasses four main areas: scientific GAI systems promise to enhance the teaching and research, public administration, business, and learning experience [10]. Taking research in higher education. education as an example, these systems are used by AI appears to have transformative potential in teachers to reduce their workload, refine instructional Italy's socio-economic landscape. However, to methods, and evaluate student performance [11]. capitalize on its benefits, high-level professional skills They can also help in creating educational resources, for developing and managing AI algorithms and such as lesson plans or worksheets, and in developing systems are essential. assessment tools like quizzes and feedback As a result, the educational strategy is primarily mechanisms [12]. On the other hand, students benefit aimed at enhancing training on AI, including GAI, from AI systems as personal tutors, helping them across educational institutions, from technical grasp difficult concepts, prepare for exams, and institutes to universities, with a particular emphasis improve their skills [13]. Moreover, scholars on PhD programs. It also includes significant recognize the potential of GAI to provide equal investments in reskilling and upskilling programs for learning opportunities, meeting the diverse needs of the workforce, ensuring that no worker is left behind all students, especially those with specific educational in this dynamic environment. These initiatives aim to needs [14]. make workers competent in using new technologies Concerns about the use of GAI systems in and to integrate AI more broadly across society to educational settings primarily stem from this mitigate long-term social and economic disparities. technology’s tendency to generate content that may Moreover, the educational strategy suggests the be inaccurate, misleading, completely erroneous introduction of AI literacy pathways in schools that (hallucinations), or that can perpetuate societal include ethical and social issues is essential, along biases, such as stereotypes or discriminatory with the establishment of internships, exchange viewpoints [15; 16]. These issues stem partly from the programs, and visiting opportunities in companies intrinsic probabilistic nature of machine learning and research institutions. technologies and partly from the potentially incorrect The summary does not explicitly mention teacher or outdated data on which these systems are trained. training. However, by addressing the issue of AI To effectively utilize GAI systems, it is essential literacy in schools, it implies that such education that the human partner demonstrates competence should be imparted to students by their teachers or by regarding the content of the conversation and can external experts. Conversely, a commitment to teacher training is moving on to some references drawn from the highlighted in the “Futura - Education for the Italian scientific literature. Future” action plan within the 2021 National The European Digital Education Hub (EDEH)’s Recovery and Resilience Plan (NRRP), which is part of Squad on Artificial Intelligence in Education, the NextGenerationEU (NGEU) programme [19]. primarily focused on primary and secondary schools, “Futura” seeks to “create a new educational system has identified three segments of teacher use of AI [24, that guarantees the right to education, digital 25, 26], effectively integrating the two dimensions competencies, and the necessary skills to face future mentioned above: challenges, thereby overcoming all forms of inequality and countering early school leaving, educational • Teaching for AI, i.e., acquiring and promoting poverty, and regional disparities” [20]. In this context, a critical view, focusing on ethics and numerous educational hubs distributed across Italy sustainability, and evaluating the offer initiatives for teachers aimed at exploring the opportunities, benefits, limitations, and potential of integrating AI and GAI in teaching and risks (e.g., biases, inaccuracies) of using AI in learning. Some “Futura” courses and workshops are society. based on the outcomes of “AI4T - Artificial Intelligence • Teaching about AI, i.e., imparting technical- for and by Teachers”, a European pilot project in computer science notions regarding AI's teacher training on AI funded by Erasmus+ and technology (e.g., Machine Learning and promoted by the European Commission as part of the Natural Language Processing). European Digital Education Action Plan 2021-2027 • Teaching with AI, i.e., employing AI systems [21]. to achieve teaching and learning goals. In the absence of an update to the Digital 4. Teacher Training on GAI Competence Framework for Educators, known as Teachers play a dual role with respect to GAI DigCompEdu [27], the Squad used existing European systems, acting both as direct users within their documentation to identify what teachers should professional practice and as facilitators in guiding know, understand, be able to do and what attitudes students in their relationship with technology. These could support them in correctly interacting with GAI roles are also related to the nature of the educational systems [25]. event, which includes an explicit curriculum, involving For the Teaching for AI segment, the Squad the formally taught content, and a hidden curriculum, employed the dimensions and examples of involving the implicit lessons learned by students as knowledge, skills, and attitudes from the last update they interact with the teacher [22]. Regarding the of the European Commission’s Digital Competence hidden curriculum, the integration of technology in Framework for Citizens, known as DigComp 2.2 [28]. the classroom by the teacher, as well as his/her For the Teaching about AI segment, the EDEH attitude and perspective, ranging from frustration and identified several competency areas based directly on skepticism to enthusiasm, may influence students’ students’ needs: basic digital skills, existing AI perceptions. A teacher who uses GAI systems in a applications, computational thinking, mathematics, proactive and visible manner and who critically and and specific AI topics. For the Teaching with AI curiously embraces new tools can encourage students segment, the Squad revisited and reorganized the to develop a positive attitude towards technology and, indicators from the European Commission’s Ethical more broadly, towards innovation [23]. This impact is guidelines on the use of artificial intelligence (AI) and further reinforced if, within the explicit curriculum, data in teaching and learning for educators [29], the teacher not only imparts knowledge and skills but aligning them with the six areas outlined in the above- also provides opportunities for discussion and mentioned DigCompEdu. constructive reflection. Regarding teacher training within the European Considering the policies of European and context, it is also worth mentioning the report on international institutions, the issue of teacher training “Professional Development for Teachers in the Age of on GAI is currently encompassed within the broader AI” from the European Schoolnet Academy [30]. topic of AI, except for a few specific documents that Without claiming to offer a framework, the report mostly propose good practices and case studies. serves as a guide for educators and policymakers on Below, we briefly analyze the proposals from how to effectively prepare teachers for a future where European bodies, the United Nations, and Italy, before AI plays a significant role in education. It emphasizes the importance of ethical considerations, continuous learning, and practical application of AI tools in that may enable or limit the use of GAI systems, such teaching, also offering examples on the topic of GAI. as state policies or educational standards [38]. This Going beyond Europe, as part of the “AI and the means that teachers need to understand not only the Futures of Learning” initiative, UNESCO is currently function of GAI and its implications for teaching and developing two new competency frameworks for AI, learning, but also its role in shaping a new ecosystem including GAI, aimed respectively at teachers and and, thus, its transformative effect on individuals and schools’ students. This framework is scheduled to be society [9]. published in September 2024 and will likely guide future AI literacy in the educational sector. The draft 5. Conclusions released in April 2024 outlines five aspects — human- centered mindset, ethics of AI, foundations and The discussion above clearly shows that the field applications of AI, AI pedagogy, and AI for professional of teacher training on GAI is receiving considerable development — and is structured around three focus from educational organizations, governmental progression levels — acquire, deepen, and create [6]. institutions, and research community. In the former In Italy, there are no specific guidelines regarding case, the focus is more shifted towards primary and teacher training on the topic of AI and GAI, which secondary education and falls within the broader could be included in the new initial training for context of AI literacy, while in the latter, the secondary school teachers established by the Decree considerations prove to be more specific and targeted of the President of the Council of Ministers dated towards the different levels of education. August 4, 2023 [31]. Authors highlight the critical importance of Scientific literature offers contributions investing in fostering a proactive and critical specifically focused on the topic of GAI and evidence approach towards GAI: teachers need to not only that can be classified with respect to different grasp how GAI works and supports teaching and educational levels. It agrees on the importance for learning, but they also need to consider their pivotal teachers to receive adequate training to develop role as educators, guiding students within a changing effective navigation skills to maximize the benefits of social ecosystem. This dual responsibility clearly GAI systems and to ensure their ethical and effective represents a challenge: teachers need to train to use use by students [32; 33; 34]. Some scholars highlight and disseminate a technology that evolves as it is the significance of teachers’ pedagogical expertise in discussed. utilizing these systems, asserting that the quality of In this regard, following a systematic review of the the output hinges on the teacher’s ability to formulate literature produced up to 2023, the research group quality questions [12; 35; 36]. In the educational authoring this paper is working on some task-based context, a prompt appears effective — i. e., experiments to obtain validated good practices. The pedagogical — if it includes variables related to a goal is to produce evidence-based guidelines for pedagogical intention, clearly refers to a disciplinary teacher training on GAI, thereby supporting the content (the subject of the topic), and is part of an educational community in critically interacting with approach where technology is considered a partner in this emerging technology. the teaching-learning process. Therefore, forging an active partnership with GAI systems seems to require References not so much digital training as rather a [1] S. Wollny, J. Schneider, D. di Mitri, J. Weidlich, M. comprehensive teacher training, in line with Rittberger, and H. Drachsler, “Are We There Yet? established frameworks like the Technological - A Systematic Literature Review on Chatbots in Pedagogical Content Knowledge (TPACK) [9; 37]. Education”, Front. Artif. Intell., vol. 4, 2021, doi: TPACK focuses on the intricate interplay between 10.3389/frai.2021.654924. technology, pedagogy, and subject matter, [2] O. Tapalova and N. Zhiyenbayeva, “Artificial empowering teachers to devise teaching strategies Intelligence in Education: AIEd for Personalised that are both effective and suitable for specific Learning Pathways”, Electron. J. e-Learning, vol. educational settings. With the introduction of GAI, one 20, no. 5, pp. 639-653, 2022. of the authors of the TPACK framework, along with [3] D. Dessì, G. Fenu, M. Marras, and D. R. Recupero, some colleagues, has started a conversation about the “Bridging learning analytics and Cognitive continued applicability and value of this proposal. The Computing for Big Data classification in micro- research group suggests that teachers must inevitably learning video collections”, Comput. Hum. give central importance to a fourth aspect, the context, i.e. must be aware of the wider systemic influences Behav., vol. 92, pp. 468-477, 2019, doi: Spec. Educ. Technol., vol. 38, no. 3, pp. 404-416, 10.1016/j.chb.2018.03.004. 2023, doi: 10.1177/01626434231165977. [4] European Commission, “European approach to [15] OpenAI, “GPT-4 Technical Report”, arXiv, 2023, artificial intelligence”, 2024. [Online]. Available: doi: 10.48550/arXiv.2303.08774. https://digital- [16] P. P. Ray, “ChatGPT: A Comprehensive Review on strategy.ec.europa.eu/en/policies/european- Background, Applications, Key Challenges, Bias, approach-artificial-intelligence. Ethics, Limitations and Future Scope”, Internet [5] UNESCO, “Artificial Intelligence”, 2024. [Online]. Things Cyber-Phys. Syst., vol. 3, pp. 121-154, Available: 2023, doi: 10.1016/j.iotcps.2023.04.003. https://www.unesco.org/en/artificial- [17] Department for Digital Transformation & intelligence. Agency for Digital Italy, “Strategia Italiana per [6] UNESCO, “AI Competency Frameworks for l’Intelligenza Artificiale 2024-2026: Executive School Students and Teachers”, 2024, Apr. 4. Summary”, Apr. 2024, [Online]. Available: [Online]. Available: https://www.notizie.ai/pathal/uploads/2024/ https://www.unesco.org/en/digital- 04/Dpd_Executive_Summary.pdf. education/ai-future-learning/competency- [18] Italian Government, “Strategic Program on frameworks. Artificial Intelligence 2022-2024”, Nov. 2021, [7] M. Bond, H. Khosravi, M. De Laat, N. Bergdahl, V. [Online]. Available: Negrea, E. Oxley, et al., “A Meta Systematic https://assets.innovazione.gov.it/1637777513- Review of Artificial Intelligence in Higher strategic-program-aiweb.pdf. Education: A Call for Increased Ethics, [19] Italian Government, “Piano Nazionale di Ripresa Collaboration, and Rigour”, Int. J. Educ. Technol. e Resilienza”, Jul. 2021, [Online]. Available: High. Educ., vol. 21, no. 4, 2024, doi: https://www.governo.it/sites/governo.it/files/ 10.1186/s41239-023-00436-z. PNRR_0.pdf. [8] C. de la Higuera and J. Iyer, “AI for Teachers, An [20] Italian Ministry of Education, “Futura. La Scuola Open Textbook”, UNESCO & European per l'Italia di Domani”, 2021, [Online]. Available: Commission, 2024. https://pnrr.istruzione.it/wp- [9] A. Iannella, “The transitional space. Generative content/uploads/2021/12/PNRR_EN.pdf. Artificial Intelligence as an opportunity for [21] European Commission, “Digital Education growth”, Ital. J. Educ. Technol., Accepted Action Plan 2021-2027: Resetting education and Manuscript Online, 2024, doi: 10.17471/2499- training for the digital age”, COM/2020/624 4324/1330. final, 2020. [Online]. Available: https://eur- [10] C.K. Lo, “What Is the Impact of ChatGPT on lex.europa.eu/legal- Education? A Rapid Review of the Literature”, content/EN/TXT/?uri=CELEX%3A52020DC06 Educ. Sci., vol. 13, 410, 2023, doi: 24. 10.3390/educsci13040410. [22] B. R. Snyder, The Hidden Curriculum, Alfred A. [11] H. Lee, “The rise of ChatGPT: Exploring its Knopf, New York, NY, 1970. potential in medical education”, Anat. Sci. Educ., [23] A. Iannella and V. Pagani, “La scuola sullo 2023, doi: 10.1002/ase.2270. schermo. Il contratto didattico e il curriculum [12] J. Jeon and S. Lee, “Large Language Models in nascosto durante la DaD e la DDI”, QTimes - J. Education: A Focus on the Complementary Educ. Technol. Soc. Stud., vol. XIV, no. 4, pp. 471– Relationship Between Human Teachers and 486, 2022. doi: 10.14668/QTimes_14433. ChatGPT”, Educ. Inf. Technol., vol. 28, pp. 15873- [24] D. Cassidy, Y.-A. Le Borgne, F. Bellas, R. 15892, 2023, doi: 10.1007/s10639-023-11834- Vuorikari, E. Rondin, M. Sharma, J. Niewint-Gori, 1. J. Gröpler, A. Gilleran, and L. Kralj, “Use Scenarios [13] T. Rasul, S. Nair, D. Kalendra, M. Robin, F. de and Practical Examples of AI Use in Education”, Oliveira Santini, W. J. Ladeira, and L. Heathcote, Eur. Digit. Educ. Hub (EDEH)’s Squad on “The role of ChatGPT in higher education: Artificial Intelligence in Education, 2023. Benefits, challenges, and future research https://www.indire.it/wp- directions”, J. Appl. Learn. Teach., vol. 6, no. 1, pp. content/uploads/2023/06/Use-Scenarios-and- 41-56, 2023, doi: 10.37074/jalt.2023.6.1.29. Practical-Examples-of-AI-Use-in-Education- [14] M. T. Marino, E. Vasquez, L. Dieker, J. Basham, 3.pdf. and J. Blackorby, “The Future of Artificial [25] Y.-A. Le Borgne, F. Bellas, D. Cassidy, R. Intelligence in Special Education Technology”, J. Vuorikari, and L. Kralj, “Teachers’ Competence”, Eur. Digit. Educ. Hub (EDEH)’s Squad on [35] I. Celik, M. Dindar, H. Muukkonen, and S. Jarvela, Artificial Intelligence in Education, 2023. “The promises and challenges of artificial Available: https://www.ai4t.eu/wp- intelligence for teachers: A systematic review of content/uploads/2023/08/AI-squad- research”, TechTrends, vol. 66, pp. 616–630, output_briefing-report-1.pdf. 2022, doi: 0.1007/s11528-022-00715-y. [26] J. Niewint-Gori, “A snapshot of the evolving [36] Y. J. Choi, H. Jang, and H. Kim, “Influence of landscape of artificial intelligence in education”, Pedagogical Beliefs and Perceived Trust on in Proc. Ital-IA 2023: 3rd National Conf. on Teachers’ Acceptance of Educational Artificial Artificial Intelligence, Pisa, Italy, May 2023. Intelligence Tools”, Int. J. Hum.-Comput. [27] C. Redecker, and Y. Punie, “European Interact., vol. 39, pp. 910–922, 2023, doi: Framework for the Digital Competence of 10.1080/10447318.2022.2049145. Educators: DigCompEdu”, Publications Office of [37] M. J. Koehler and P. Mishra, “What Happens the European Union, Luxembourg, 2017, doi: When Teachers Design Educational Technology? 10.2760/159770. The Development of Technological Pedagogical [28] R. Vuorikari, S. Kluzer, and Y. Punie, “DigComp Content Knowledge”, J. Educ. Comput. Res., vol. 2.2: The Digital Competence Framework for 32, pp. 131–152, 2005, doi: 10.2190/0EW7- Citizens”, Publications Office of the European 01WB-BKHL-QDYV. Union, 2022, doi: 10.2760/115376. [38] P. Mishra, M. Warr, and R. Islam, “TPACK in the [29] European Commission, “Ethical Guidelines on age of ChatGPT and Generative AI”, J. Digit. the Use of Artificial Intelligence and Data in Learn. Teach. Educ., vol. 39, pp. 235-251, 2023, Teaching and Learning for Educators”, doi: 10.1080/21532974.2023.2247480. Publications Office of the European Union, 2022, doi: 10.2766/153756. [30] M. Cukurova, L. Kralj, B. Hertz, and E. Saltidou, “Professional Development for Teachers in the Age of AI”, European Schoolnet, Brussels, Belgium, 2024. [31] Decree of the President of the Council of Ministers, “Definizione del percorso universitario e accademico di formazione iniziale dei docenti delle scuole secondarie di primo e secondo grado, ai fini del rispetto degli obiettivi del Piano nazionale di ripresa e resilienza”, 4 Aug. 2023, (23A05274), [Online]. Available: https://www.gazzettaufficiale.it/eli/gu/2023/ 09/25/224/sg/pdf. [32] R. Peres, M. Schreier, D. Schweidel, A. Sorescu, “On ChatGPT and beyond: How generative artificial intelligence may affect research, teaching, and practice”, Int. J. Res. Mark., vol. 40, no. 2, pp. 269-275, 2023, doi: 10.1016/j.ijresmar.2023.03.001. [33] A. N. Ansari, S. Ahmad, and S. M. Bhutta, “Mapping the Global Evidence Around the Use of ChatGPT in Higher Education: A Systematic Scoping Review”, Educ. Inf. Technol., 2023, pp. 1–41, doi: 10.1007/s10639-023-12223-4. [34] G. van den Berg, and E. du Plessis, “ChatGPT and Generative AI: Possibilities for Its Contribution to Lesson Planning, Critical Thinking and Openness in Teacher Education”, Educ. Sci., vol. 13, 998, 2023, doi: 10.3390/educsci13100998.