Generative artificial intelligence (GenAI) is rapidly reshaping higher education by automating many tasks historically used to develop and assess student competence, from drafting essays and summarizing literature to writing code and answering exam-style questions. This chapter examines GenAI’s implications for universities through a combined technical, pedagogical, and institutional lens. It first explains the foundations of large language models, emphasizing their probabilistic nature and the resulting limits, including hallucinations and unreliable factuality, which make verification and critical judgment essential academic skills. It then analyzes the opportunity–risk landscape for higher education: potential gains in productivity for educators, increased access to always-available tutoring and cognitive support for students, and new possibilities for ideation and feedback, alongside risks of dependency, erosion of cognitive effort, academic misconduct that is difficult to detect reliably, and privacy and security concerns. To move from reactive prohibition to principled integration, the chapter presents an intervention model implemented at Iscte – University Institute of Lisbon, combining structured GenAI literacy initiatives for faculty and a foundational course for students. The chapter concludes with institutional recommendations for assessment redesign, curriculum integration of AI literacy, and policy approaches that preserve academic integrity while enabling responsible experimentation, arguing that GenAI ultimately forces a re-clarification of what higher education values in learning.

--

• Ciências da Computação e da Informação - Ciências Naturais
• Ciências da Educação - Ciências Sociais