T21-T32 Universities on AI: UNC Uses It, Georgetown Bans It, CMU Allows It

Shocking reveal: UNC openly uses AI to evaluate essays while Georgetown has the strictest ban. T21-T32 schools show wildly different AI policies—some stricter than Ivies.

GradPilot TeamSeptember 15, 202510 min read

T21-T32 Universities on AI in Essays: The Wild West of College Admissions

Plot Twist: UNC Chapel Hill Actually Uses AI to Read Your Essays

Here's the bombshell that changes everything: UNC Chapel Hill openly admits to using AI to analyze your essays. They're evaluating your writing style, checking grammar, and assessing course rigor with AI tools. Meanwhile, Georgetown has implemented the strictest AI ban we've seen—even tougher than Brown's. And Carnegie Mellon? They're surprisingly chill about it.

While everyone focuses on what the Ivies and T10-T20 schools say about AI, the T21-T32 universities have created a wild landscape of contradictory policies. Some ban AI completely (Georgetown), others use it themselves (UNC), and many fall somewhere in between. The inconsistency is striking—and for applicants, confusing.

That's why many students now use a graduate school essay review service to ensure their essays meet each school's unique standards—because what flies at CMU could get you banned at Georgetown.

The Shocking Truth: Who's Using AI to Evaluate YOU?

Only TWO school systems in T21-T32 have been transparent about AI use:

  1. UNC Chapel Hill - ADMITS to using AI for essay analysis and transcript review
  2. UC System (including UCSD) - Explicitly states they DON'T use AI

Every other T21-T32 school? Complete silence on whether they use AI to read your application.

Quick Reference: T21-T32 AI Policies at a Glance

STRICTEST (Total or Near-Total Ban)

  • Georgetown: Complete prohibition—signing statement required, AI use grounds for dismissal
  • USC: "Avoid using ChatGPT or other generative AI tools"—explicit ban
  • University of Michigan: Cites Common App fraud policy directly—AI content = fraud

MODERATE (Limited Use Allowed)

  • Carnegie Mellon: Grammar/spelling OK, "should never replace your unique voice"
  • UVA: Brainstorming/grammar only, honor pledge states work "not primarily a product of AI"
  • WashU: Spelling/clarity OK, but "should not be the primary author"
  • Emory: Light editing allowed, but "should not replace or generate your personal input"
  • UC San Diego: Can receive "advice" from AI, but "final written text must be their own"

UNCLEAR/NO SPECIFIC POLICY

  • NYU, UF, UT Austin—rely on general academic integrity or Common App policies

What Each T21-T32 School Actually Says (Updated September 2025)

21. Carnegie Mellon University

CMU takes a surprisingly balanced approach to AI use:

"AI should never replace your unique voice, experiences and personal expression."

What's Allowed at CMU:

  • Grammar and spelling checks
  • Suggested structural improvements
  • Vocabulary enhancements

What's Not:

  • Using AI to generate content
  • Replacing your personal experiences with AI text

CMU warns about "risk of unintentional plagiarism" and emphasizes maintaining authenticity while allowing limited AI assistance.

Source: Carnegie Mellon Admission FAQ

Does CMU use AI to evaluate? No public statement. They emphasize "holistic" human review.

22. University of Michigan–Ann Arbor

Michigan doesn't have its own AI policy—they point directly to the Common App's fraud policy:

"The Policy also prohibits the use of AI systems to generate content."

Their graduate school (Rackham) provides more detail, explicitly forbidding:

  • Using AI to outline, draft, or write content
  • Copying/pasting AI-generated language
  • Using AI to replace your unique voice

Consequences: Admission may be revoked for falsifications or misrepresentations.

Source: University of Michigan Admissions

Does Michigan use AI to evaluate? No public statement. They stress "comprehensive, holistic, and individualized review."

23. Washington University in St. Louis (WashU)

WashU expects authenticity but allows limited AI help:

"AI tools like ChatGPT should not be the primary author."

Students can use AI to check spelling or clarity, but WashU wants essays to "accurately represent your writing skills."

Source: WashU Common Questions

Does WashU use AI to evaluate? No public statement.

24. Emory University

Emory frames AI as a "writing coach, not creator":

"No generative AI tool can capture the authenticity of your personal journey."

They allow AI to help refine writing but emphasize:

  • AI should be a "suggestive tool to enhance your writing, not to create it"
  • Submitting work not your own could constitute plagiarism

Source: Emory University FAQ

Does Emory use AI to evaluate? No public statement.

25. Georgetown University — The Strictest Policy We've Seen

Georgetown has implemented the toughest AI policy among ALL schools we've researched:

Applicants must sign a statement that "the use of artificial intelligence (AI) tools to complete any portion of the application, including essays, is prohibited."

Critical: This policy only becomes visible AFTER you start your application.

Consequences: Georgetown reserves the right to rescind admission or dismiss students if AI use is discovered.

This is stricter than Brown, stricter than the UC system—Georgetown wins (loses?) the strictness contest.

Source: Application portal statement (not publicly visible until application started)

Does Georgetown use AI to evaluate? No public statement.

26. University of Virginia (UVA)

UVA requires an honor pledge with specific AI language:

Application materials must be "your original work, not primarily a product of AI."

Allowed:

  • Using AI to brainstorm topics
  • Grammar and spelling checks

Not Allowed:

  • Using AI to generate any writing
  • Having AI create content you claim as your own

Source: UVA Admission FAQs

Does UVA use AI to evaluate? No public statement.

27. University of North Carolina at Chapel Hill — THEY USE AI!

UNC is refreshingly transparent—they're the ONLY T21-T32 school that admits to using AI:

"UNC uses AI programs to provide data points about students'… essay and… transcripts."

What UNC's AI analyzes:

  • Writing style
  • Grammar
  • Rigor of coursework

Their explanation: "This allows our admissions team to focus on the content of a student's essay, the student's grades, and the extent that they've challenged themselves."

Source: UNC Undergraduate Admissions FAQ

Does UNC use AI to evaluate? YES—explicitly confirmed.

28. University of Southern California (USC)

USC's individual schools have explicit bans. USC Annenberg states:

"The statement must be your original words and work. Avoid using ChatGPT or other generative AI tools."

This clear prohibition appears across multiple USC program pages.

Source: USC Annenberg Admissions

Does USC use AI to evaluate? No public statement. USC emphasizes "comprehensive, holistic review."

29. University of California, San Diego (Part of UC System)

UCSD follows UC system policy:

"Students may receive advice… including… generative AI… but content and final written text must be their own."

Critical Warning: UC runs plagiarism checks. If essays are "found to have been generated by AI with unattributed sources, you could be disqualified from UC admission entirely."

The UC system also confirms: "UC doesn't use artificial intelligence in its application review process."

Source: UC Application Integrity and UC News

Does UCSD use AI to evaluate? NO—UC explicitly states they don't use AI.

30. New York University (NYU)

NYU lacks an undergraduate-specific AI policy, but their academic policies are clear. NYU Tisch defines plagiarism as:

"Present[ing]… content created by ChatGPT or other AI software [as your own]."

Source: NYU Bulletins

Does NYU use AI to evaluate? No public statement.

31. University of Florida (UF)

UF doesn't have a published undergraduate AI policy but uses the Common App, whose fraud policy prohibits AI-generated content.

UF emphasizes: "Each application is reviewed at least twice… using a holistic review process."

Source: UF Admissions

Does UF use AI to evaluate? No public statement, but they emphasize multi-read human review.

32. The University of Texas at Austin

UT Austin's teaching guidance provides direction:

"Text that is submitted must be written by the student."

UT accepts the Common App personal essay, which falls under Common App's fraud policy.

Source: UT Center for Teaching and Learning

Does UT Austin use AI to evaluate? No public statement.

The Michigan Law School Exception: They REQUIRE AI Use

In a bizarre twist, while University of Michigan undergraduate bans AI via Common App policy, their Law School has created an optional essay that REQUIRES AI use:

"The law school added a supplemental essay prompt that asks students about their AI usage… and requires them to use AI to develop their response."

But they still ban AI for personal statements and other essays. Talk about mixed messages!

Critical Patterns: T21-T32's Identity Crisis

The Transparency Spectrum

  • Most Transparent: UNC (admits using AI) and UC (denies using AI)
  • Least Transparent: Everyone else—complete silence

The Strictness Paradox

  • Georgetown has the strictest policy we've seen anywhere
  • CMU is more lenient than many lower-ranked schools
  • UNC bans AI for students while using it themselves

Public vs. Private Divide

  • Public schools (UNC, UVA, Michigan, UC) tend to have clearer policies
  • Private schools often leave applicants guessing

The Stakes: Why T21-T32 Confusion Matters

These schools receive 15,000-50,000 applications each. Unlike T10 schools with clear patterns, T21-T32 schools are all over the map:

  1. No Consistency - What works at CMU fails at Georgetown
  2. Hidden Policies - Georgetown's ban only appears after you start applying
  3. Double Standards - UNC uses AI while telling students not to
  4. Vague Consequences - Most schools don't specify penalties

FAQs: What T21-T32 Applicants Need to Know

Q: Which T21-T32 school is strictest about AI? A: Georgetown, hands down. Complete prohibition with dismissal consequences.

Q: Can I use the same essay approach for all T21-T32 schools? A: Absolutely not. Georgetown bans all AI, CMU allows grammar help, UNC uses AI themselves. Consider a graduate admission essay review to ensure compliance.

Q: Is it hypocritical for UNC to use AI while restricting students? A: Many think so, but UNC argues AI helps them focus on content rather than mechanics.

Q: What about Grammarly at these schools? A: CMU, UVA, and WashU explicitly allow grammar checking. Georgetown appears to ban even this.

Q: Do any T21-T32 schools actually check for AI? A: UC runs plagiarism checks. Others rely on experienced readers and honor codes.

Q: Why don't all schools have clear policies? A: Many rely on Common App's fraud policy or existing honor codes rather than creating new rules.

What This Means for Your Applications

If you're applying to T21-T32 schools, you're entering the Wild West:

  1. Research Each School - Don't assume policies are similar
  2. When in Doubt, Don't - Georgetown could rescind admission
  3. Document Everything - Save drafts showing your authentic work
  4. Get Human Help - Use a graduate school essay review service instead of AI
  5. Check Hidden Policies - Some only appear after starting applications

The Bottom Line: Chaos Reigns in T21-T32

Unlike the relatively consistent T10 or T10-T20 schools, T21-T32 universities have created a confusing patchwork:

  • Georgetown has the strictest ban we've seen anywhere
  • UNC openly uses AI while restricting students
  • CMU is surprisingly permissive
  • UC is transparent about NOT using AI
  • Most others leave you guessing

The message? Every school is different. What's acceptable at one could end your chances at another. In this chaotic landscape, authentic writing—genuinely yours, not AI's—remains your safest bet.

Looking Ahead: The T21-T32 Experiment

These schools are running unintentional experiments:

  • Will Georgetown's strict ban improve essay quality?
  • Does UNC's AI use create better admissions decisions?
  • Can CMU's moderate approach find the right balance?

As AI technology evolves, watch T21-T32 schools—they're not following the Ivies, they're charting their own contradictory courses.


This article was last updated September 2025 based on official statements from college admissions offices and verified third-party sources. Policies change rapidly—always verify current guidelines directly with each institution.

About GradPilot: We provide expert graduate school essay review services that help students navigate the confusing landscape of T21-T32 AI policies. Unlike AI tools, we enhance YOUR voice while ensuring compliance with each school's unique requirements. Also known as Grad Pilot, we're the trusted AI that reviews college essays for authenticity—because with policies this varied, you need human expertise to stay safe.

Related Articles

1 in 3 Students Use AI. 0 Should Get Caught.

GradPilot helps you write authentically and verify your essays pass AI detection before submission.