Students Use AI to Prove Their Essays Are Human

Students Are Humanizing Their Writing—By Putting It Through AI

Teachers and students are caught in a tug-of-war: educators rely on AI detection tools to catch cheating, while students use the same technology to prove their work is genuine. The result? Students run their own writing through AI detectors to make sure it doesn’t get flagged as robotic or machine-generated.

Miles Pulvers, a 21-year-old student at Northeastern University, never uses AI to write essays. Yet, he runs every paper through an AI detector before submitting it. “I take great pride in my writing,” he says. “Before AI, I had peace of mind that whatever I submitted would be accepted. Now some of my writing gets flagged as AI-generated when it’s not. It’s annoying, but that’s how it is in 2025.”

AI detectors often flag writing that includes many adjectives, long sentences, or certain punctuation like em dashes. When this happens, Pulvers rewrites the flagged sections and tests again until the detector shows a low chance of AI involvement. This process is becoming common as teachers grow increasingly suspicious, often for good reason.

“They’re using AI to write their essays, and then they use AI detectors to humanize them,” says Leticia Villaseñor, a history teacher at a private high school in Los Angeles. She relies on AI detection software to catch cheaters and prefers in-class written assignments because “if it’s homework, half of them will use AI.”

I’m Not a Robot

Proving authenticity has become a challenge for students who write their own essays. Marcus Wooler, a 16-year-old student, shared that some of his classmates faced wrongful accusations and had to prove their work was original by showing Google Doc histories. To avoid this, Wooler runs his essays through AI detectors before submitting them. “It’s a good precaution,” he says.

Using formal language or grammar-checking tools can ironically increase the chance of being flagged. Devan Leos, co-founder of Undetectable AI, points out that words like “delve” or “valuable insight” can raise red flags. “It’s like you get penalized for being a proper writer,” he explains.

Harrison Checketts, 16, took a different approach after running his English paper through five AI detectors with inconsistent results. Some tools flagged parts as AI-written, others did not. He decided to trust his teachers, who “are the real AI detectors” because they can recognize a student’s voice.

Claire Krieger, a 20-year-old Fordham University student, also tests all her papers with AI detectors but warns they can create a false sense of security. One detector flagged quotes she cited, while another flagged a sentence she wrote herself. She resolved the issue by meeting her professor and affirming her commitment to academic integrity.

A ‘Police State’ of Writing

The consequences for students accused of submitting AI-generated work are serious. One University of North Georgia student was placed on academic probation after being accused of plagiarism for using Grammarly to correct grammar—not to cheat. The incident pushed Grammarly to develop an authorship tool that tracks the writing process and flags AI usage, generating reports students can share with teachers.

Jenny Maxwell, Grammarly’s head of education, says the fear around AI in schools is causing students to spend extra time rewriting to sound more human. Over 500,000 people use Grammarly’s AI and plagiarism detection tools each week, most of them students. She describes the current environment as a “police state of writing.”

For writers, educators, or anyone involved in content creation, understanding how AI detection tools work—and how they influence writing styles—is crucial. If you want to learn more about AI tools and how they impact writing and creativity, check out Complete AI Training’s latest AI courses.