The tool should feel like a demanding tutor, not a ghostwriter
Across Brazil, students facing the vestibular — one of the most demanding college entrance exams in the world — are turning to artificial intelligence not as a shortcut, but as a new kind of mirror, one that reflects their arguments back to them at midnight with tireless precision. The technology promises to democratize access to high-quality writing feedback, reaching students whose families cannot afford private tutors. Yet educators remind us that the deepest purpose of the essay has always been to reveal how a mind works under pressure — and that a mind shaped entirely by algorithmic suggestion may not be a mind that has truly learned to think.
- Brazilian vestibular students, especially those from lower-income families, are embracing AI writing tools that offer instant, personalized feedback once available only to those who could afford elite prep courses.
- The tension is not simply about cheating — it runs deeper: educators fear students may learn to follow algorithmic corrections without ever developing the metacognitive awareness that makes a writer genuinely capable.
- Specialists are drawing a careful line, urging students to use AI for generating prompts and spotting structural flaws, but insisting that the hard intellectual work — choosing arguments, weighing evidence, deciding what to keep — must remain human.
- Universities and institutions are now confronting an urgent policy gap, needing to define what responsible AI use looks like in exam preparation before the technology outpaces any meaningful oversight.
- The vestibular itself faces an identity question: if AI has polished every essay submitted to it, can it still measure the quality of thought it was designed to reveal?
Brazilian students preparing for the vestibular — the high-stakes college entrance exam that opens or closes the doors to the country's most competitive universities — are increasingly turning to artificial intelligence to practice their essay writing. These platforms offer something traditional tutoring rarely can: immediate, personalized feedback on argumentation, structure, and clarity, available at any hour and at a cost far below that of a private instructor.
The vestibular essay is a formidable challenge. Students must build persuasive arguments on unfamiliar topics under strict time pressure, demonstrating not just grammar but the capacity to think critically and organize complex ideas on demand. AI tools have stepped into this space by generating exam-style prompts, analyzing submissions in real time, and flagging logical or structural weaknesses before they become entrenched habits. For students without access to expensive prep courses, this represents a genuine democratization of test preparation.
But educators are urging caution. Their concern goes beyond the risk of cheating — it touches something more fundamental. If students outsource the diagnosis of their own writing problems to an algorithm, they may never develop the metacognitive skills that separate strong thinkers from those who simply follow suggestions. The vestibular essay is meant to reveal how a student thinks; if AI has smoothed every rough edge, what exactly is being measured?
The practical guidance emerging from specialists is clear: use AI for what it does well — practice prompts, structural feedback, alternative phrasings — but do the hard thinking yourself. Argue with the feedback. Decide what to accept and what to reject. Understand why a sentence works. The tool should function as a demanding tutor, not a ghostwriter.
As AI adoption accelerates across Brazilian education, institutions face a reckoning. The goal is not to ban these tools, but to establish guidelines that harness their strengths without hollowing out the learning process. The vestibular will evolve — the open question is whether it will continue to measure what it was always meant to measure.
Brazilian students preparing for the vestibular—the grueling college entrance exam that determines admission to the country's most competitive universities—are turning to artificial intelligence to sharpen their essay writing. The tools offer something traditional tutoring often cannot: immediate, personalized feedback on structure, argumentation, and clarity, available at any hour and at a fraction of the cost of a private instructor.
The vestibular essay is notoriously demanding. Students must construct persuasive arguments on unfamiliar topics within strict time limits, demonstrating not just grammatical competence but the ability to think critically and organize complex ideas under pressure. AI platforms have begun filling a gap in preparation resources by generating writing prompts tailored to exam specifications, analyzing student submissions in real time, and flagging weaknesses in logic or organization before bad habits calcify.
For many students, particularly those from families without access to expensive preparatory courses, these tools represent a democratization of test prep. A student can submit an essay at midnight, receive detailed commentary on their thesis statement, evidence selection, and conclusion, and revise immediately. The feedback is consistent, tireless, and free from the judgment or fatigue that can color human instruction. Some platforms even simulate the timed conditions of the actual exam, allowing students to practice under realistic constraints.
But educators are sounding a careful note. The consensus among vestibular prep specialists is clear: AI should amplify human instruction, not replace it. The concern is not merely about cheating—though that risk exists—but about something subtler and more corrosive. If students come to rely on AI to diagnose their writing problems, they may never develop the metacognitive skills that distinguish strong writers from those who simply follow algorithmic suggestions. Critical thinking, the ability to recognize when an argument is weak or when evidence doesn't support a claim, cannot be outsourced to a machine without cost.
There is also the question of authenticity. The vestibular essay is meant to reveal how a student thinks. If AI has shaped every sentence, smoothed every rough edge, and optimized every transition, what exactly is being measured? Educators worry that widespread AI use could create a generation of students who can produce polished essays but cannot think their way through a difficult problem without algorithmic assistance.
The practical advice emerging from specialists is measured. Use AI for what it does well: generating practice prompts, identifying structural problems, suggesting alternative phrasings. But do the hard thinking yourself. Argue with the feedback. Decide which suggestions to accept and which to reject. Understand why a sentence works or fails. The tool should feel like a demanding tutor who knows your weaknesses and won't let you slide—not a ghostwriter.
As AI adoption in Brazilian education accelerates, institutions face a reckoning. Universities will need to establish clear guidelines about what constitutes acceptable use of these tools in exam preparation, and perhaps in the exams themselves. The goal is not to ban the technology—that would be both futile and counterproductive—but to harness it in ways that strengthen rather than hollow out the learning process. The vestibular will evolve. The question is whether it will measure what it was designed to measure.
Citas Notables
Use AI for generating practice prompts and identifying structural problems, but do the hard thinking yourself— Vestibular preparation specialists
La Conversación del Hearth Otra perspectiva de la historia
Why are Brazilian students turning to AI for vestibular prep specifically? What makes this exam different from other college entrance tests?
The vestibular is brutal and high-stakes. It determines which university you attend, which shapes your entire career trajectory. There's no second chance, no portfolio, no demonstrated interest over time. It's one essay, one day, and the pressure creates demand for any tool that promises to help.
But couldn't a student just hire a tutor?
A tutor costs money many families don't have. AI is free or cheap. That's part of why educators see it as democratizing—but also why they're nervous. If only wealthy students learn to think critically and poor students learn to follow AI suggestions, the inequality gets worse, not better.
What's the actual risk if a student relies too heavily on AI feedback?
They stop developing their own judgment. They learn to recognize when AI says something is wrong, but not why it's wrong. They become dependent on external validation instead of trusting their own reasoning. That's fine for an essay, but it's a dangerous habit for a person.
So the tool itself isn't the problem?
No. The tool is neutral. The problem is how it's used. AI as a mirror—showing you what you wrote and asking you to think harder—that's valuable. AI as a crutch, doing the thinking for you, that's corrosive.
What would responsible use look like?
Write first. Then use AI to identify problems. Then sit with those problems and decide what to do. Argue with the feedback. Reject suggestions that don't feel true to what you're trying to say. The friction between your thinking and the machine's suggestions—that's where learning happens.