LISBON, Jan. 20 (Xinhua) -- In a quiet university library in central Portugal, laptop screens glow late into the evening. Some students hesitantly type, while others scroll, copy, and paste. One question spreads faster than any new software update: is artificial intelligence (AI) expanding minds, or draining them?
That question now sits at the center of a growing debate in Portuguese higher education. Dozens of professors from universities and polytechnic institutes across the country have signed a manifesto calling for a ban on generative AI in teaching and learning, warning that students are being turned into "digital cretins."
The manifesto, signed by 28 academics, urges institutions to "promote the humanization of higher education" and to suspend the use of generative AI tools in academic processes. Students, the text argues, are the main victims of a digital environment dominated by large language models and chatbots that encourage plagiarism, intellectual laziness, and what the authors describe as a "factory of banalities."
"The health of students is collapsing, anxiety levels are soaring, and, transformed into digital cretins, they show little intellectual curiosity or enthusiasm for the demanding adventure of knowledge," the manifesto states.
Behind the provocative language lies a deeper concern. The professors argue that the rapid and largely uncritical adoption of AI by higher education institutions - driven by fear of "missing the train of progress" - has outpaced serious reflection on pedagogy, ethics, and cognitive development. In their view, conferences, workshops, and working groups have replaced firm decisions, while classrooms remain largely unchanged.
The discussion focuses on a practical challenge: how should education adapt when technology can produce answers faster than students can think?
These dilemmas in Portugal reflect broader international concerns. The Organization for Economic Co-operation and Development (OECD) this week published its Digital Education Outlook 2026, which examines global research on generative AI in education. One cited study involved students from five American universities who were asked to write a short essay. Those who used AI tools received higher grades, but only 12 percent of these students could accurately quote their own text one hour later. Among students who used no AI or only search engines, 89 percent could do so.
The OECD described this gap as "metacognitive laziness," a term now frequently cited by critics of AI in education.
Still, the debate is far from one-sided. Portugal's Minister of Education Fernando Alexandre has rejected the idea of simply turning back the clock. Speaking during a visit to schools this week, he said AI "is a reality that cannot be ignored."
"What we must do is adapt - through teacher training, changes in teaching methods and curricula - to maximize the opportunities and minimize the risks," he said, arguing that AI should become a complementary tool to enhance students' capacities.
This tension between prohibition and adaptation divides educators themselves. Joseph Wheeler, who has been a professor for 20 years and was interviewed for this report, says skepticism is understandable, but warns against denial. "As educators, we need to embrace it, not ignore it," he said. "Misuse can be detrimental to learning, and studies clearly show that students who rely on AI remember very little. But there are guardrails that can be put in place."
Wheeler sees vast potential if AI is used carefully. "It can adapt to individual students, stimulate learning, and ease teachers' workloads," he said, while stressing that technology cannot replace the human core of education. "There is no substitute for a good teacher."
That view is shared by Gilberto Franca, a consultant developer and technology enthusiast, who considers that the progress of AI is irreversible. "It will be integrated so naturally into daily life that even people with difficulties using technology will be using it without realizing it," he said. The danger, Franca argues, lies in shallow learning. "When students just copy and paste, they don't create connections with the information, and become dependent on AI."
For Franca, the solution depends largely on educators. "If teachers use these tools well and make learning more dynamic, both students and teachers will grow," he said, pointing to improved accessibility and faster access to information.
Students themselves occupy the uneasy middle ground. Breno Moreira, an engineering student in Portugal, says that AI is already an integral part of academic life. "It's a path with no return," he said. "But it has to be used consciously. It shouldn't do the work for us."
Moreira describes AI as especially helpful in theoretical subjects, allowing students to study at their own pace and clarify doubts that crowded classrooms cannot always address. Nevertheless, he worries about over-reliance. "In an exam, you won't have an AI with you. You'll only have what you actually learned," he said.
He believes institutions should respond with education, not bans. "There should be training on AI throughout the course, not just at the end," he said, suggesting the creation of institution-specific educational AI systems to reduce risks and misinformation.
Back in the faculty offices, the authors of the manifesto remain unconvinced. They argue that without a pause - a suspension - universities risk accelerating a cognitive impoverishment that will be difficult to reverse. For them, the issue is not technology itself, but the speed and scale of its intrusion into learning.
As Portugal's universities weigh their next steps, the debate reflects a broader global dilemma. AI promises efficiency, personalization, and access. It also challenges memory, effort, and authorship - foundations on which education has long rested.
Between the glowing screens in libraries and the warnings written in manifestos, one thing is clear: the future of learning will not be decided by algorithms alone, but by how societies choose to live with them.
