IN OUR rush to keep up with Artificial Intelligence, schools have started using “AI detectors” to check if students are writing their own work. These tools are supposed to protect academic honesty. But in reality, they are creating a new kind of injustice: they often mistake real Filipino writing for machine‑made text.
A recent study among Filipino college students shows how serious the problem is. Using two popular detectors—ZeroGPT and Copyleaks—half of the student essays that were written by hand were wrongly flagged as AI‑generated. Imagine that: a 50 percent chance of being accused of cheating even when you did the work yourself.
The reason is simple. These detectors are trained mostly on native English writing. When they encounter “Filipino English,” which may use simpler words, direct sentences, or different grammar patterns, they confuse it with the style of AI. For Filipino students, writing clearly should be a strength, not a reason to be suspected.
And this problem is not only for undergraduates. Graduate students, doctoral candidates, and even professors can be affected. A thesis, a research paper, or a journal article could be wrongly labeled “AI‑generated.” That kind of mistake can damage reputations, block scholarships, or even stop a career.
In the Philippines, education is the most trusted path out of poverty. To be falsely accused of dishonesty because of a flawed algorithm is devastating. It can mean failing grades, lost opportunities, and broken trust between teachers and students.
What makes things worse is that these detectors are inconsistent. One tool may say “Human,” while another says “AI” for the same essay. If the machines themselves cannot agree, how can schools use them as proof?
Writing in a second language is already hard. Adding the fear of being “profiled by an algorithm” makes students anxious instead of confident. It sends the wrong message: that their natural way of writing in English is “not human enough.”
The answer is to bring back the human element. AI detectors should only be used as a first check, never as final evidence. Teachers and schools must rely on human judgment, review drafts, and talk openly with students. Technology should help learning, not replace fairness.
As we enter this new age of AI, we must make sure innovation does not erase human voices. A student’s essay, a master’s thesis, or a professor’s research paper is a human achievement. No algorithm should have the power to take that away.
Learning is not a code to be scanned—it is a voice to be heard.
——————————————————————————————
Mhel Cedric D. Bendo is a student researcher and opinion writer at the Polytechnic University of the Philippines, specializing in Educational Psychology, Curriculum and Instruction, and Educational Technology. He has authored multiple peer‑reviewed articles.
