Learning AI #23

Is it cheating to have AI write your college essay?

Should AI help write your child’s college essay?

Ever since I accidentally burnt holes in my pajamas after experimenting with a chemistry set on my eighth birthday, I have always had a passion for science.

This statement was copied word-for-word in more than 5% of 50,000 medical school essays sampled. Something was wrong before AI hit the scene.

But is it cheating to let AI write your kid's college essay?

Yes, if the AI writes it. No, if the AI helps your kid write it.

College essays exist to do one thing: show an admissions officer how a specific person thinks, not how well that person can prompt a chatbot. 

If a parent feeds a kid's report card and list of extracurriculars into ChatGPT and pastes the output into the Common App, the school is admitting a fiction.

The student didn't write 650 words about resilience. The AI did.

But that's not the only way people use AI on these essays, and most of the realistic uses aren't cheating.

A kid writes a rough draft about caring for a younger sibling during a parent's illness. He runs it through Claude and asks where it's vague and which paragraph is weakest. That's called editing.

Writers have always had editors. A teacher who circles a paragraph and writes "expand this" is doing the same job an AI does when it flags that the fact that the second paragraph repeats many of the statements made in the first.

Brainstorming works the same way. A student who can't find a topic and prompts an AI to ask him a series of questions about important events in his life hasn't outsourced anything. He still has to write the answers.

The test isn't whether AI touched the essay. It's whether the finished piece reflects the kid's own thinking, in the kid's own voice, about something true.

Admissions officers have read the writings of thousands of 17-year-olds. They can tell when a sentence sounds like a 45-year-old executive wrote it.

A polished essay that sounds like nobody is a bigger fail than a rough essay that sounds like someone.

A parent working his child’s college essay through AI is different than the child conversing with AI and brainstorming ideas.

There's a longer-term cost parents aren't weighing. The essay is one of the only places in the entire application process where a teenager has to sit with a blank page and figure out what they believe about themselves.

Skip that, and the kid arrives on campus a few months later having never done it. A professor could ask the same kind of question on a first assignment. The kid doesn’t have the skills he should have learned during the application process. Fail.

AI for feedback, structure, and pushing back on weak reasoning is fine. AI for generating the actual sentences that get submitted is not, and most selective schools' policies already say so directly.

The schools that ask “tell us about a time you failed" aren't grading the prose. They're using the essay as the one part of the application a parent's money and a tutor's polish can't fully reach. The moment a parent runs the whole thing through AI, that bit of uniqueness, or the essence of what makes the kid who he is, gets erased.

If your kid is using AI to find out what they think, let them. If you're using it to skip the part where they figure that out, you've done their homework for them, and the school never finds out until your kid struggles through his first midterms or accidentally burns a hole in his pajamas.

Things I think about

Fingernails grow four times faster than toenails.

Free gift for my subscribers

Here is the link for a free audiobook chapter from my new book, No Adult Left Behind. AI is Learning. Are You?

There is no login or anything like that, just click and listen. The chapter is all about how to use prompts with AI. A good starting point for anyone who is serious about learning AI. (The audiobook is narrated by my AI-cloned voice.)

Response from ChatGPT

Usually: yes, if the AI writes the essay and the student presents it as their own work. But there’s nuance.

College essays are generally trying to answer:

Can this student think, reflect, communicate, and tell their own story?

If AI generates the substance and voice, many admissions officers would view that as misrepresentation, even if there isn’t a formal “AI detector.”

A useful way to think about it:

Generally acceptable

  • Brainstorming topics

  • Asking AI for questions to reflect on

  • Getting help organizing structure

  • Editing for clarity, grammar, or length

  • Getting feedback (“Does this sound repetitive?”)

Risky / often crosses the line

  • AI writes the first full draft

  • AI invents stories or emotions

  • AI rewrites until it no longer sounds like the student

  • Parent + AI produce the final essay and the student lightly edits

Usually considered cheating or dishonest

  • Paste prompt → generate essay → submit

  • Submit AI-written content as authentic personal reflection

Admissions readers actually care less about perfect writing than people think. A distinctive, slightly imperfect essay often reads as more credible than a polished AI-sounding one.

A guideline I like:

AI should act like a coach or editor, not a ghostwriter.

If the student can honestly say, “These are my experiences, my ideas, and mostly my words—AI helped me sharpen them,” they’re probably on solid ground.

For a college essay specifically, I’d aim for:

  • Student generates stories and outline

  • Student writes draft

  • AI helps tighten and question assumptions

  • Parent helps with judgment, not authorship

That tends to produce better essays anyway.

Response from Gemini

Parents absolutely should help their kids understand how to navigate AI during the college essay process, but they need to exercise strict boundaries.

The short answer is: AI can be a partner for brainstorming and polishing, but it should never be used to outline, draft, or write the essay.

The college admissions landscape enforces a zero-tolerance approach to AI-generated drafts, and misusing these tools can severely jeopardize a student's chances.

🚫 The Red Lines: Where AI Will Get a Student Rejected

  • Generating Full Drafts or Paragraphs: Having ChatGPT or Claude write a "rough draft" that the student intends to edit later is a massive mistake. AI text is heavily formulaic, prone to trite clichés ("Through this journey, I learned..."), and lacks the granular, sensory details of a real human experience.

  • The "AI Voice" is Obvious: Admissions officers read thousands of essays; they easily spot the sterile, structurally perfect, but completely soulless "AI voice." Furthermore, the Common App officially classifies copying and pasting AI-generated text as application fraud.

  • University Penalties and Detectors: Most major university systems use advanced AI detection software. Institutions like the University of California system explicitly state that applications found using unattributed AI-generated content can be disqualified entirely.

The Green Light: Ethical and Effective Ways Parents Can Guide AI Use

Instead of banning AI entirely, parents should teach their kids how to use it responsibly as a collaborative coach rather than a ghostwriter.

1. The Socratic Brainstormer

Staring at a blank page is the hardest part of the process. Instead of telling AI to write an essay, the student can use it to interview them.

Great Prompt: "I want to write my college essay about my experience rebuilding computers for low-income students. Act as an interviewer and ask me 5 probing questions one by one to help me uncover the deeper personal meaning behind this activity. Do not write any part of the essay for me."

2. Feedback on Flow and Structure

Once a student has written a messy first draft entirely on their own, AI is great for checking structural integrity without altering the actual words.

Great Prompt: "Read this draft of my college essay. Do not rewrite anything. Instead, give me bullet-point feedback on whether my transitions make sense and if the conclusion feels strong."

3. Final Copyediting and Grammar Checking

Using AI at the very end to catch typos, passive voice, or misplaced commas is widely accepted. In fact, schools like Caltech and Cornell explicitly permit the use of basic editing tools (like Grammarly) as long as the ideas and voice remain entirely the student's.

Great Prompt: "Highlight any grammatical errors or awkward phrasing in this paragraph, but do not change the vocabulary or tone."

💡 The Ultimate Rule of Thumb

If you or your child are wondering if a specific use of AI is ethical, Caltech Admissions suggests a simple test: Ask yourself whether it would be ethical to have a high school teacher or parent do the exact same task.

Is it okay for a parent to proofread a completed draft for typos? Yes. Is it okay for a parent to write the outline or draft the paragraphs? Absolutely not. Treat AI with those exact same boundaries.

**********