Stephen Onakuse, from the Department of Food, Business and Development at Cork University Business School, discusses the potential risks of AI overuse in third-level education, and outlines the considerations we need to make to use AI responsibly in educational settings.
Third-level institutions pride themselves on independent thinking and the way they engage students with information on how to read, write, think, remember, and solve problems, which quietly shapes the structure of students’ minds. These processes are deeply interconnected, through the organised undergraduate, MSc, and PhD training programmes that form a cognitive ecosystem that determines not just what we know, but how we understand the world.
Artificial intelligence (AI) is changing how we read, write, think, remember, and solve problems. Reading is often treated as passive absorption, but it is an active reconstruction of meaning. The rise of AI in teaching and learning, however, poses the risk of reduced cognitive effort, with the potential to confuse AI learning speed with thinking. The real question is whether AI is becoming a substitute for cognition in teaching, learning, and research, and the kind of deductive thinking third-level education needs to continue to keep students’ minds capable and stable.
Efficiency or laziness?
The book Anxious Generation by Jonathan Haidt is leading the charge worldwide on our collective moral obligation, starting with a suggested total ban on phones in schools in the AI Supercycle. The book explores how the decline of free-play in childhood and the rise of smartphone use among adolescents is changing our world.
The phone-addicted generations are now in the habit of cognitive offloading – a slow replacement for teaching, learning, and research efforts that keep attention, memory, and judgment sharp. Cognitive offloading risk is not theoretical, and it’s already measurable in how students approach teaching, learning, and research problems. The instinct to prompt before thinking is becoming a reflex that’s worth watching.
The more AI takes over teaching, learning, and research in third-level education, the more invaluable undisciplined thinking becomes. Capability increases, but so does the need to decide when not to use it. Disciplined thinking is what third-level education should be practicing. AI in education can make it so much easier to outsource our thinking, creating the potential for our brains to become lazy.
Risks of AI in education and how we can combat them
Third-level education is faced with the problem of developing mental fitness trainers to complement the risks associated with conflating cognition with standardised tests, which are also not compatible historically with teaching, learning, and research.
The use of AI in education can be a danger to students’ cognitive learning, especially when they have not had enough time or experience to learn many of the skills that AI can easily replace. One of the main challenges is how students will learn and adapt to the changing world with AI.
Third-level education across the globe continues to deal with a system designed in the mid-19th century that is supposed to educate students, ground them in cross-functional skills, and prepare them for a world that no longer exists and will not exist in the future when they graduate from college. Today, students using AI can only experience the outcome, but not the process behind it. From our phones to computers, we can now generate images, essays and videos in seconds, solve complex problems instantly, and automate tasks that once required entire group/team efforts.
When a student cannot do basic mathematics or remember and apply basic facts to broader contexts, it becomes a systemic problem that is being either exposed or denied by technology. Many have argued that we need to consider digital metacognition to develop healthy habits around devices and AI, to sync with how universities’ programmes aligned with AI. Protecting time for deep thinking may become one of the most important teaching and learning habits third-level institutions need to insist on.
The worry of AI in third-level education is how easily it can short-circuit teaching and learning efforts, by not protecting the time for deep thinking, making AI become students’ de facto. Education generally encourages students to think, find solutions, remember the right order of a process, and understand how to use deductive thinking through a complex system. Teaching, learning, and research equip and train students’ brains in such a way that they can improve every day.
Protecting cognitive function
Today, students tend to reach for AI tools first, rather than doing the hard thinking beforehand. The real skill is knowing. Independent thinking allows for reflection and better decision-making. The habits we develop through teaching, learning, and research determine the depth of our cognition. Therefore, the cognitive alarm bell is really going to become more important for the younger generations in many years to come, to manage, support, govern, and ensure that it supports humanity. For this to happen, cognitive abilities to assess, question, understand and respond are essential. Accepting AI hype, without understanding the holistic implications, continues to create issues in areas that will impact most of humanity.
Human memory is not just storage, it is the foundation of understanding. Strong memory allows us to recognise patterns, draw connections, and solve problems efficiently. Convenience in AI use should not replace the mental effort needed for deep thinking and reasoning. Learning techniques – such as spaced repetition, active recall, and meaningful association – strengthen memory, making learning more durable and usable.
Skills acquired through education enable students to develop skilled problem-solving apparatus as they organise knowledge better and approach challenges with structured thinking. The distinction between efficiency and actual understanding is becoming more critical. Maintaining cognitive effort will be key in an AI-driven world.
The real risk with AI is not that it replaces work, it’s that it replaces the effort that builds thinking ability. In the words of Dr Jared Cooney Horvath, “this is the first generation where kids are not only not smarter than parents, but they are significantly dumber.” Therefore, thinking FOR self, by SELF, with SELF is a diminished ability. In as much as AI makes cognition effortless, it does not build memory and judgment, which is a key concern of losing outsourcing thinking to speed.
Holistic education is very important, but weakening any one layer of teaching, learning, and research eventually affects the whole system, even if it’s not obvious immediately. Speed from tools is easily mistaken for depth of thought, leading to shallow cognitive engagement over time. When tools begin to think with us or for us, the risk isn’t immediate failure, it’s gradual erosion of attention, memory, and independent reasoning through repeated substitution.
In many third-level institutions, chronic class lecture absenteeism has become increasingly common. This is particularly the case following the COVID-19 pandemic, when digital lecture material and class recordings became a lot more available. If a large share of students are missing sustained exposure to structured learning, effortful thinking, or at least basic school practice, attention, memory, and reasoning will diminish. Students develop their brains through a myriad of intricate processes that continue to strengthen synaptic connections. Critical thinking and meta cognitive processes are the seat of creative and critical thinking. Over-reliance on tools such as AI may reduce problem-solving ability over time. Formal classroom education keeps thinking skills active and strong through different scenarios.
How can we use AI in education responsibly?
Efficiency always comes with a hidden cost. When the shortcut gets too good, it stops being a shortcut and becomes the destination. Today, AI is the shortcut to many destinations. Technology is heavily embedded within education. Should third-level institutions allow AI to replace the cognitive effort that builds judgement, memory, and reasoning over time through teaching learning and research? Used well, it can enhance thinking, but it still needs to be balanced with moments where we do the mental work ourselves.
The rise of AI has pushed third-level education to outsource and make the effort of teachers much harder to notice when we’ve started outsourcing the thinking to AI. AI has changed these processes dramatically. Constant notifications fragment attention, search engines reduce reliance on memory, and short-form content encourages surface-level engagement. While these tools increase access to information, they can weaken deep cognitive skills if used uncreatively.
Our cognitive offloading improves efficiency but can reduce skill retention over time. These abilities are not fixed. They are shaped by how we use them on a daily basis. By becoming more intentional about how students study, write, think, remember, and solve problems, we can strengthen the very architecture of our minds. Speed from AI can feel like progress, but real understanding still needs human effort and attention. Speed without thinking is still reckless, especially when it does not equal clarity. The important thing now is what must be done to start fixing this paradox. Lawmakers, educators, professionals, parents, and students themselves must come together and reverse this trend. Otherwise, AI will shape how we think, create, and solve real-world challenges.
Please note, this article will also appear in the 27th edition of our quarterly publication.
