Abstract:

This contribution presents the results of the first phase of the EU-funded HITS project (High-Intelligent Tutoring System – 2023-1-IT02-KA220-HED-000152340), which aims to design a flexible, ethical, and student-centred AI-based tutoring platform for higher education. The project's initial focus was to understand the real needs, expectations, and challenges related to AI-supported tutoring from multiple perspectives, students, educators, tutors, and academic experts, through an extensive Participatory Action Research (PAR) process.

The methodology combined desk research with over 370 stakeholder engagements, including surveys and semi-structured interviews conducted across 57 European Higher Education Institutions (HEIs). This phase led to the co-creation of a Needs and Prioritisation Map, highlighting key educational priorities such as personalization, ethical AI integration, early identification of student difficulties, improved guidance, and support for digital inclusion.

Findings confirm that while stakeholders acknowledge the potential of AI in transforming tutoring and academic guidance, concerns remain about data privacy, transparency, and the perceived risk of replacing human educators. Many students showed interest in AI-assisted systems that adapt to their learning paths and offer real-time feedback, yet they stressed the importance of maintaining a human dimension in learning. Educators, in turn, emphasized the need for training, clear ethical standards, and tools that complement rather than replace their role.

The results are currently informing the development of a modular, scalable tutoring platform that blends AI with human mentorship. This includes defining system requirements and functionalities such as adaptive learning, continuous assessment, early-warning systems for dropout prevention, and ethical monitoring tools. The project also produced a draft Ethical Code of Conduct for AI in Higher Education, co-designed with academic partners and external experts, now available for consultation.

This presentation will outline the participatory methodology used, summarize the main outcomes, and reflect on their implications for the design and implementation of intelligent tutoring systems. It will also open critical discussion on how universities can effectively and ethically adopt AI technologies to support both academic performance and inclusion.

To stimulate dialogue with the academic community, the paper concludes with a set of open questions:
- How can institutions ensure AI supports, rather than replaces, human interaction in learning?
- What role should educators, senior students, and experts play within AI-enhanced tutoring systems?
- What behavioural and cultural changes are needed among students and staff to foster meaningful AI adoption?
- How can universities balance innovation with ethical responsibilities when deploying AI tools?

Addressing these questions, the HITS project contributes to shaping an inclusive, transparent, and pedagogically sound future for AI in higher education.

Keywords:

Ethical AI integration, Intelligent tutoring systems, Higher education.