The availability to the public of the Generative Artificial Intelligence tool ChatGPT has led to several reactions in society at different levels. Regarding higher education several challenges have arisen, especially in terms of ethics and evaluation, and its integration into teaching and research practices. In this study, we intend to explore mainly the issues related to the integration and ways of using ChatGPT in higher education, especially in initial teacher training, and the implications of this use for education policies and global citizenship.

With the rapid development and widespread accessibility of Generative Artificial Intelligence (Gen-AI), it is paramount to understand its implications in various areas of society, in terms of knowledge creation and its contribution to the Sustainable Development Goals (UNESCO, 2021), notwithstanding the necessary epistemological reflection on its use (Figueiredo, 2023).

In higher education, Artificial Intelligence (AI) has the potential to completely transform teaching and learning (Rawas, 2023). The potential of ChatGPT shows remarkable benefits in teaching, research support, automated grading, administrative management, and human-computer interaction (Dempere et al., 2023). It can provide individualized recommendations to students, increase collaboration and communication, and further improve their learning outcomes (Rawas, 2023). However, have been identified ethical concerns and implementation issues about security in student assessment and plagiarism, misuse, and the possibility of misinformation, as well as wider social and economic impacts such as job displacement, the digital literacy gap, or decreased human interaction (Dempere et al., 2023; Rawas, 2023).

ChatGPT, as a Gen-AI tool, can help conversationally with writing, learning, solving and assessment, as an assistant for instructors and a virtual tutor for students (Lo, 2023). A literature review highlights measures relating to assessment methods and the necessary institutional policies. Rethinking assessment tasks to reduce the risk of plagiarism by requiring students to demonstrate their skills in real-time and in person, for example. Course content, learning outcomes and assessment methods can also be modified to circumvent ChatGPT, by using it to generate lesson topics, test and exam questions, homework, or product ideas (De Winter, 2023).

On the other hand, from a more constructive and training perspective, it will also be important to promote students' digital literacy in the use of Gen-AI tools. Teaching students about the risks of relying on AI-based technologies is important. These risks include hallucinations, which are false responses generated by AI, presented as facts, not explained by the training data (Dempere et al., 2023). For this reason, it is important to integrate these technologies responsibly, as a supplement to and not a replacement for human interaction (Fuchs, 2023), and there is a pressing need to regulate AI in Higher Education Institutions (HEIs).

As far as initial teacher training (ITT) is concerned, this phenomenon is even more relevant, since these students, as future teachers, will soon be training pupils in education systems. It requires teachers and students develop digital competences and literacies, with a strong focus on critical thinking and fact-checking strategies (Kasneci et al., 2023).

A qualitative approach will be used with recourse to non-participant observation and narrative research methods through the analysis of experiences developed in the curricular unit Initiation to Professional Practice of a Master’s in Teaching. To this end, data was collected taking into account: i) what are the main difficulties and constraints in use; ii) what are the benefits in the planning and preparation of classes; iii) what are the adaptations to instructional methods, form of assessment, and pedagogical practices needed to use the ChatGPT in the teaching and learning process in an ethical and safe way. In addition to the data from the empirical study, supported by the literature review, two Gen-AI tools, ChatGPT and Elicit, were trialled and their outputs analysed. Given the recent availability of these Gen-AI tools to the public, quality scientific studies published in the Scopus and WoC databases on this subject are still scarce, and the quality of the articles mobilised was prioritised over quantity. The study's qualitative approach took a naturalistic and hermeneutic perspective, using content analysis of the field notes from non-participant observation and of student narratives carried out as a final assignment (Amado, & Freire, 204; Bardin, 2013). This methodology is often used in research in the social sciences and education, as the researcher is dealing with complex situations in which it is difficult to select variables. In this way, the researcher seeks to describe and analyse a phenomenon and its interactions and does not intend to quantify or generalise. The narrative research method provides in-depth knowledge of the respondents' experiences and is based on a constructivist and interpretive epistemology (Rabelo, 2011). It considers that a narrative can express the complexity of the experience, as well as the relationships and uniqueness of each action (Bolívar et al., 1998), allowing knowledge to be obtained through an account that captures the details of meanings beyond factual statements or abstract propositions. Finally, it should be noted that informed consent was obtained from the study participants, thirteen preservice teachers, and their identity and anonymity were safeguarded, in accordance with the institution's ethics charter and international benchmarks, as Ethical Guidelines for Educational Research (BERA, 2011).

Generative AI literacy will be an indispensable asset, as it provides students with the skills to critically engage with AI systems, ensuring that they become active and discerning users. At the same time, prompt engineering makes it possible to improve the outputs generated in a more precise way and enables educators and students to maximize the usefulness of the educational resources created by AI (Bozkurt, 2023). This study corroborates that, for the development of AI literacy, it is important to acquire proficiency in understanding, interacting with and critically evaluating generative AI technologies, which is essential not only for the current digital age, but also for shaping the future of education. It is also important to understand the ethical considerations, prejudices and limitations inherent in such systems, as well as to promote critical thinking and digital citizenship among students, teachers and researchers. So, Gen-AI literacy can and should be integrated into the curriculum to cultivate a new generation of informed and responsible users, and teachers should adapt their teaching methods to incorporate AI, preparing students for a future where it is an integral part of their personal and professional lives. The impact of AI on education and higher education cannot be ignored, and it is essential to integrate it into teacher education as well (Moura, & Carvalho, 2024). Recommendations include emphasizing a humanistic approach, mobilizing interdisciplinary planning, empowering teachers, and enhancing trust and safety. It also concludes that it is essential to address and include issues relating to artificial intelligence in higher education and to reflect them in legislation and educational policy.

Amado, J., & Freire, I. (2014). Estudo de caso na Investigação em Educação [Case study in Education Research]. In Manual de investigação qualitativa em educação [Handbook of qualitative research in education], (pp.121–168). Imprensa da Universidade de Coimbra. Bardin, L. (2013). Análise de Conteúdo [Content Analysis]. Edições 70. Bolívar, A., Domingo, J., & Fernández, M. (1998). La investigación biográfico–narrativa en educación. Guía para indagar en el campo. [Biographical-narrative research in education. A guide to research in the field.]. Grupo FORCE, Universidad de Granada, Grupo Editorial Universitario. Bozkurt, A. (2023). Unleashing the Potential of Generative AI, Conversational Agents and Chatbots in Educational Praxis: A Systematic Review and Bibliometric Analysis of GenAI in Education. OpenPraxis, 15(4), 261–270. https://doi.org/10.55982/openpraxis.15.4.609 De Winter, J.C.F., Dodou, D., & Stienen, A.H.A. (2023). ChatGPT in Education: Empowering Educators through Methods for Recognition and Assessment. Informatics, 10, 87. https://doi.org/10.3390/ informatics10040087 Dempere, J., Modugu, K., Hesham, A., & Ramasamy, L.K. (2023). The impact of ChatGPT on higher education. Front. Educ., 8, 1206936. https://doi.org/10.3389/feduc.2023.1206936 Ethical Guidelines for Educational Research (BERA) (2011). Available online: https://eera-ecer.de/about-eera/ethical-guidelines (accessed on 9th January 2024). Figueiredo, A. D. (2023). Inteligência Artificial Generativa e Construção de Conhecimento (Generative Artificial Intelligence and Knowledge Building). Personal communication. In Processamento de Linguagem Natural: Tendências e Aplicações Práticas Conference. https://doi.org/ 10.13140/RG.2.2.25801.52328 Fuchs, K. (2023). Exploring the opportunities and challenges of NLP models in higher education: is Chat GPT a blessing or a curse. Front. Educ, 8. https://doi.org/10.3389/feduc.2023.1166682 Kasneci, E., Sessler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., et al. (2023). ChatGPT for good? On opportunities and challenges of large language models for education. Learning and Individual Differences, 103, 102274. https://doi.org/10.1016/j.lindif.2023.102274 Lo, C.K. (2023). What Is the Impact of ChatGPT on Education? A Rapid Review of the Literature. Educ. Sci., 13, 410. https://doi.org/10.3390/educsci13040410 Moura, A., & Carvalho, A. A. (2024). Literacia de Prompts para Potenciar o Uso da Inteligência Artificial na Educação [Prompt Literacy to Enhance the use of Artificial Intelligence in Education]. RE@D - Revista de Educação a Distância e Elearning, 6(2), e202308. https://doi.org/10.34627/redvol6iss2e202308 Rawas, S. (2023). ChatGPT: Empowering lifelong learning in the digital age of higher education. Educ Inf Technol. https://doi.org/10.1007/s10639-023-12114-8 Rabelo, A. O. (2011). A importância da investigação narrativa na educação [The importance of narrative enquiry in education.]. Educação & Sociedade, 32(114), 171-188. https://doi.org/10.1590/S0101-73302011000100011 UNESCO (2021). AI and education: Guidance for policy-makers. UNESCO. https://doi.org/10.54675/PCSP7350