A sweeping Mexico survey shows generative AI is already embedded in higher education, reshaping writing, reasoning, and even emotional support. At the same time, universities scramble to regulate it, train teachers, and protect critical thinking before convenience hardens into nationwide academic dependence.
A Classroom Revolution Already Happened
In Mexico, the debate over generative artificial intelligence in higher education has already moved beyond theory. The technology is not arriving. It has arrived, unpacked its bags, and taken a seat in the classroom.
That is the clearest message running through the national survey on the uses and perceptions of generative AI in Mexican higher education, cited by Wired and conducted by the Secretariat of Public Education. More than 60% of students and professors in higher education use generative AI systems in their daily routines. In a country where public debate often lags behind technological adoption, that number lands with force. It suggests Mexican universities are not facing a distant disruption but living inside one already.
The scale of the study reinforces that point. More than 1.5 million students and 166,000 teachers from public and private universities participated, a figure the notes describe as the broadest survey of its kind in the world. The findings show overwhelming familiarity with generative AI. Ninety-three percent of students and ninety-five percent of teachers say they know about it. Sixty-nine percent of students and seventy-three percent of teachers say these tools have improved academic performance. Those are not numbers of marginal experimentation. There are several normalizations.
And the uses are not narrow. Eight out of ten students use systems such as ChatGPT or Gemini to generate academic texts. Image creation accounts for 61% of mentions, followed by code generation at 15%. Yet the most revealing figures may be those that go beyond technical convenience. Seventy-nine percent of teachers and eighty-two percent of students say these tools are useful for complex cognitive processes such as reasoning, reflection, imagination, and creativity. That is a far more consequential claim than simple productivity. The university community in Mexico is beginning to treat generative AI not just as an assistant, but as a participant in thought itself.
This is where the story becomes larger than campus technology. In Mexico, higher education has long been asked to do several jobs at once. It must produce professional mobility, national development, public legitimacy, and intellectual formation, often under conditions of inequality and uneven infrastructure. When generative AI enters that ecosystem so quickly and so deeply, it does not just change classroom habits; it changes the way we live. It begins to change the texture of what academic formation means.
The Convenience Trap
The survey, as presented in Wired’s reporting, does not read as either utopian or alarmist. It reads as uneasy. It acknowledges the usefulness of generative AI while warning that its spread is unfolding amid weak training, low confidence, and unclear rules.
That combination may be the most important part of the whole story.
Despite widespread usage, self-perceived mastery of the technology is low, averaging 5 on a 10-point scale. Ninety-one percent of teachers and 76.2% of students say they need training. Meanwhile, 76.7% of teachers and 67.2% of students in public institutions say they do not know the institutional rules governing the use of these tools. In other words, Mexico’s universities have embraced AI faster than they have learned to govern it.
That is not just an administrative issue. It is both pedagogical and cultural. A campus can quickly adopt a powerful tool because the incentives are obvious. Writing becomes faster. Research feels more manageable. Drafting, coding, and brainstorming become less solitary. But when regulation and formation trail so far behind, convenience starts to organize the learning environment before institutions have defined what they actually want students to learn from it.
Mario Delgado, head of the Secretariat of Public Education, frames the challenge in exactly those terms. According to the notes, he argues that the discussion should not focus on permitting or prohibiting these technologies in classrooms, but on defining their pedagogical uses, setting clear rules, and delimiting their scope. That is a serious position, and probably the right one. Prohibition was never likely to work once use became this common. But neither is surrender. What Mexico appears to need is not panic, but a new educational grammar for a reality that has outrun the old one.
Carlos Iván Moreno Arellano says it even more bluntly. The current teaching and evaluation models, he suggests, may no longer make sense in an era when generative AI advances much faster than norms, public policy, and regulations. “At this moment, we are failing,” he warns. That sentence cuts through the euphemisms. The problem is not merely that universities are late. It is that the existing academic architecture may be poorly designed for a world where machines can draft, summarize, imitate, and assist at the level students are now experiencing daily.
There is a special tension here in Mexico, where debates about modernization often carry the weight of both sovereignty and dependence. The country seeks technological strength, digital competitiveness, and trained talent to support a changing economy. But it also knows the risks of importing systems faster than it can shape them for national needs. Generative AI sharpens that old dilemma. It promises efficiency and inclusion, yet it can also weaken critical habits if adopted passively.
What a Mexican University Is For
One of the most striking details in the survey is that nearly 92,000 students and just under 6,000 teachers say they use these systems for emotional support, whether to cope with anxiety episodes or simply as a way to vent. That figure changes the moral register of the conversation.
This is no longer only about essays, images, or code. It is about the emotional economy of university life in Mexico.
A student who turns to AI for comfort is not just looking for a faster answer. That student may be looking for patience, availability, or relief in an environment where institutional support can feel thin. This does not automatically make AI harmful. But it does show how quickly a tool designed around language can slide into a role once reserved for teachers, friends, counselors, or family. The question for Mexico is not only whether these systems can help students perform. It is whether universities are prepared for what happens when performance, thought, and emotional dependence begin to blur.
That is why the ten principles proposed by the Secretariat matter. The notes describe them as an ethical, pedagogical, and inclusive framework. They call for recognizing generative AI as a consolidated academic tool, establishing clear institutional guidelines, prioritizing teacher training and digital literacy, updating curricula, rethinking assessment, guaranteeing student literacy in these systems, reducing access and governance gaps, incorporating gender perspectives, addressing student well-being, and strengthening the humanities and social sciences across university formation.
That last point may be the most important of all. Andrés Morales, UNESCO’s representative in Mexico, says the debate over generative AI is not only technological, but also ethical, philosophical, human rights-based, and pedagogical. He adds that digital transformation only makes sense if technology serves humanity and not the other way around. In a moment like this, that sounds less like a slogan than a warning.
Mexico’s universities are being asked to choose what kind of intelligence they want to form. If generative AI becomes just another shortcut in an already unequal system, then students may graduate faster into a thinner version of knowledge. But if institutions treat this moment seriously, with rules, literacy, critical thinking, and stronger humanistic formation, then the country may yet turn an unruly technological wave into something more durable.
The real issue, then, is not whether Mexico will use AI in higher education. It already does. The issue is whether its universities will teach students to command the tool without quietly surrendering the work of judgment, doubt, and thought that a university is supposed to protect. Wired’s reporting makes clear that this future is already being negotiated in Mexican classrooms. The question is whether the institutions meant to guide that negotiation can move quickly enough to matter.
Also Read:
Mexico Watches Surveillance Boom Blur Safety, Secrecy, and the Border
