Artificial Intelligence (AI) is increasingly integrated into higher education as a tool to personalize learning, support autonomous study, and enhance teaching efficiency. However, the current reality of AI in education is far from the dystopian imaginaries of popular culture. This paper analyzes best practices for the pedagogical use of AI in higher education, highlighting strategies such as guided prompt design, ethical awareness, and reflective activities that prevent passive dependence on automated responses. Methodologically, the research adopts a reflective and practice-oriented approach implemented throughout a semester in a core course of the Bachelor’s Degree in Industrial Technologies Engineering. AI-driven tools—including chatbots (e.g., ChatGPT, DeepSeek…), generative content platforms, and adaptive quiz systems—were integrated into a Blended-Learning framework that combined asynchronous tasks via Moodle with synchronous face-to-face sessions. The Flipped-Teaching model was used to deliver conceptual content prior to class through short video capsules and AI-generated summaries. During in-person sessions, students participated in structured activities such as guided prompt design workshops, critical analysis of AI-generated responses, and group discussions on ethical implications and algorithmic bias. Each week, students completed AI-based formative assessments followed by peer review sessions, encouraging the comparison between human- and AI-generated outputs. These activities were designed to enhance critical thinking, evaluative judgment, and ethical reflection while promoting active, student-centered learning.
Beyond the initial opportunities offered by AI in higher education, its effective use requires the development of specific good practices that ensure a pedagogical balance between support and autonomy. Among these, prompt design emerges as a critical skill, as the way students formulate instructions directly influences the quality, accuracy, and relevance of AI-generated responses. Encouraging learners to engage in prompt engineering promotes not only better outputs but also a reflective process of questioning, refining, and iterating. Similarly, AI-based tutoring systems can complement traditional teaching by offering personalized guidance, instant clarification, and formative feedback. Ethical considerations are also essential, including academic integrity, avoiding plagiarism, appropriate use and citation of AI tools, and understanding the boundaries between assistance and substitution.
The results highlight that AI can meaningfully complement active learning methodologies, strengthening digital competencies, and promote reflective and engaged student participation. Students reported improvements in their ability to design purposeful prompts, assess the reliability of AI outputs, and recognize limitations such as algorithmic bias or lack of contextual depth. Furthermore, the integration of AI-based tutoring systems provided opportunities for immediate feedback and autonomous exploration, while still maintaining the guidance of the instructor as a key mediator. Students positively valued the different academic activities developed during the experience, as well as the opportunity to explore AI-tools within a structured and ethically guided learning environment, highlighting their potential to enhance learning processes while fostering autonomy, analytical reasoning, and responsible use of emerging technologies.

Keywords: Artificial Intelligence AI, Chatbots, Virtual assistant, Critical Thinking Training, Generative and Conversational AI tools, Flipped-Teaching, AI-based activities, Blended Learning, Game-based Learning and Gamification, New experiences in education, New Trends and Experiences.