AI and the disruption of higher education

The exponential development of generative AI capabilities is already having a significant impact on the education sector. 

Just over two years ago, shortly after the launch of ChatGPT (January 2023), Khan Academy introduced Khanmigo, a personalized tutor that guides students through virtually every subject in primary and secondary education, and even beyond. Khanmigo has since evolved and now operates in over 570 school districts across the United States, used daily by nearly 100,000 students and teachers. Its second version is about to be released, featuring a much more interactive approach—proactively asking questions and tracking students’ progress. 

Tech giants are also entering the education space with their own solutions. The most notable is Google’s Gemini, which extends its Google Classroom offering. It’s designed to support teachers by helping them create tutors (similar to OpenAI’s GPTs), quizzes, syllabi, presentations, exams, and more. Seamlessly integrated into Google Classroom, it offers a comprehensive teaching support environment. This is not about reinventing education, but about equipping it with modern AI tools that make it more effective. 

But AI-based tutors have gone even further. A recent study conducted in Nigeria showed that six months with an AI tutor were equivalent to two full years of academic progress. 

Meanwhile, Stanford psychology graduates MacKenzie Price and Andrew Sutherland founded a network of AI-driven schools, including Alpha School and Alpha High School. They claim that thanks to these tutors, students learn two to three times faster, requiring only two hours a day for academic subjects. The rest of the day is spent on projects, group activities, music, and sports. 

The results of many of these AI-driven schools support their model, placing them in the top 2% in Texas and Arizona

Many of these schools operate as charter schools—a legal figure in the US that enables the creation of innovative education projects. Most are located in Texas, but there are also schools in Arizona and Toronto. While some are relatively affordable, most charge around $40,000 per year per student—albeit with impressive facilities. 

The results of many of these schools—though not all—support their model, placing them in the top 2% of educational institutions in Texas and Arizona. These states, let us recall, are home to renowned universities like Texas A&M and Arizona State University (ASU), considered one of the most innovative in the US. 

And it’s not just these schools. Online platforms like Duolingo, Speak, or Jumpspeak have revolutionized language learning with apps that correct pronunciation, allow speaking from day one, and assess progress instantly. Today, language instruction that doesn’t include AI already feels like a relic of the past. 

A new educational paradigm

Few doubt that AI tutors will play a central role in education. Their presence in classrooms will likely become indispensable for younger students, freeing up time for what truly defines education—what goes beyond the mere transmission of knowledge. 

Their use in higher education is also just a matter of time—and not much of it. This is especially true in executive education and in highly structured subjects like algebra, calculus, economics, Python, and many others. 

Some even predict the emergence of two-year degree programs. In countries deeply concerned with educational performance—China, South Korea, Taiwan, and Japan—AI is already being tested through pilots at all levels. These are countries where efficiency is a key value, including in education. 

Universities have not remained untouched. One of the most emblematic cases is Harvard’s CS50 course (Introduction to Computer Science), led by David J. Malan. Likely Harvard’s most popular course, it is offered both in-person and online, with over six million learners. The in-person version alone has 1,000 students and more than 2,000 hackathon participants. To manage this, it used to rely on over 100 teaching assistants—until the course team developed its own AI tutor to replace them. This is just one of many pilots currently underway at universities around the world. 

Still inconclusive results

We do not yet have rigorous studies on the effectiveness of these pilots: for now, the results are more anecdotal than definitive, though quite promising. It’s true that Alpha School attracts families from the tech sector, with high cultural capital and deep concern for their children’s future—students who would likely succeed in any environment. The data therefore lacks statistical strength. Still, the idea that personalized, continuously evaluated learning is more effective is difficult to dispute. 

We are witnessing the most significant disruption in the history of education

These tutors enable experiences far beyond what was previously possible: talking to Julius Caesar, interrogating Einstein or Winston Churchill, engaging in deep Socratic dialogues, or learning calculus progressively at each student’s own pace. Upcoming advances in virtual reality will take these experiences to unforeseen levels—spending a day in imperial Rome, navigating the human body or distant galaxies, or participating in business simulations of unprecedented realism. 

In addition, tools like Google’s Veo 3, which generate videos from text descriptions, make it possible to build these experiences without massive investment. They are now within reach for many. 

The Disruption

We are witnessing the most significant disruption in the history of education. It’s now possible to deliver instruction that is two or three times more effective, on a massive scale—virtually unlimited—and at a low and decreasing cost. 

The economic disruption is clear, but so is the disruption to the teaching profession itself. Content delivery and 90-minute lectures (or 3-hour ones in business schools) are being replaced by personalized sessions with AI assistants. The focus is shifting toward experiential learning and project-based work, with an unprecedented level of realism. 

We will see these possibilities explored in countless directions. Some will compress degree programs into two years; others will offer bachelor’s-plus-master’s combinations in just four years. 

As with all innovation, the most disruptive changes will arise at the margins. Nonconformists and outliers will lead the charge. But countries obsessed with educational performance—China, Korea, Japan—are already launching ambitious programs to incorporate these technologies into their educational systems. 

Online education will become a key space for the most disruptive proposals

In business schools, the first major shift will occur in executive education, which is poised to undergo the most significant transformation in its history. But let’s not forget: executive training is at the core of a business school’s mission. It won’t be long before these technologies reach MBA programs and other advanced degrees. 

However, adoption will vary widely. Some schools will embrace highly disruptive formats, making AI tutors central to the learning experience. Others will opt for a more incremental approach, equipping faculty and students with generative AI tools to boost effectiveness. At the same time, online education will become a key space for the most disruptive proposals. 

Educational institutions are heavily regulated and often lack the flexibility to choose their pace of adoption. Yet technological progress will inevitably push them toward substantial changes. In this context, there are only three strategic options: 

• The first is to do very little—adopt only the minimum, what everyone already knows (ChatGPT, etc.). Taking this stance means aligning with the most conservative players, which in a time of disruption entails taking a significant risk.
• The second strategy is to adopt best practices—that is, to keep pace with the leading institutions. No one gains a major competitive advantage this way, but no one falls behind either. It’s a conservative strategy, but one that allows for an accelerated pace of change if necessary.
• The third strategy is to rethink education more profoundly—developing pilots that go beyond individual subjects and operate at the program level, enabling the acquisition of substantial competitive advantages and using the disruption to move forward rather than merely stay afloat. Some will make technology the core of their development; others will seek to reach new heights in the holistic development of the human being, knowing that, ultimately, this is the very purpose of education. 

Conclusion: beyond the technological promise

Generative AI seems destined to revolutionize almost everything, including professions that have remained largely unchanged for centuries—like teaching. 

To this day, teaching still resembles what it has been for generations: storytelling, explaining. Once done with little to no tools, then with blackboards, later with overhead projectors, and now with sophisticated PowerPoint slides. Ultimately, it’s always about using imagination to transport a class to past worlds or convey abstractions that explain how societies and organizations work. 

AI allows to return to early models of personalized education and extend them to large populations

Modern education began during the Renaissance. Humanism introduced more scientific pedagogy and sought to extend education to broader segments of the population. The Jesuits created the first international network of schools and universities, expanding across Europe and beyond. From Esade, we are part of that legacy—and now we have the duty and the opportunity to update it. 

On one hand, AI allows us to return to early models of personalized education through tutors. On the other, it makes it possible to extend this personalized approach to large populations, without sacrificing quality. 

The technological promise is compelling: a personal tutor who tirelessly repeats concepts using fresh examples and engaging exercises, who only focuses on what the student has yet to learn, and who can help review ideas, test hypotheses, and turn them into examples through fictional experiments. 

But is that enough to raise the overall level of education and reach higher degrees of intellectual and human development? 

Perhaps not enough attention is being paid to this question. Our experience with students shows that a side effect of such powerful knowledge acquisition systems is that they devalue knowledge itself. Why invest effort in learning something that isn’t immediately useful? 

Moreover, the impressive pace of technological progress is occurring alongside a growing devaluation of the labor market. Never before have so many professions felt as threatened as they do today by AI-driven disruption. 

So the promise of rapid learning is not enough. We must reinforce the why of learning. We must explore other dimensions of knowledge that go beyond the optimized interaction between a student and a machine. 

The real challenge lies in integrating technological progress with the goal of delivering more holistic, meaningful education

Finally, we should reflect on the value system behind these technologies. The celebrated emphasis on efficiency can breed dependence and increase competitiveness among students. As is already happening among adults, greater access to knowledge doesn’t lead to limits in academic dedication—it simply raises the bar in competing with other universities and institutions. 

In our view, the real challenge lies in integrating technological progress with the goal of delivering more holistic, meaningful education. But the balance of power is uneven. Technology arrives with force and urgency. The educational mission requires stronger awareness and the building of a committed learning community. This is a precious and vital task that lies ahead. 

Technological change, transformation, and disruption do not stem from technology itself—they are driven by people with ambition that goes beyond themselves. Managing these individuals and teams is what sets leading institutions apart from mediocre ones—and Esade must be among the former.