Should ChatGPT Write Your College Application Essay?

Applying to college can be stressful, and many students claim the personal essay causes the most anxiety along the way.

What if there were a way to simplify the essay process, a shortcut easing the burden of putting pen to paper, metaphorically speaking, and drafting ideas to tell your story?

Well, we all know that shortcut exists in the form of artificial intelligence (AI), most notably ChatGPT and its mind-boggling ability to crank out copy on demand in record time.

But will using AI yield the best results? Is it even ethical? And can admissions officers recognize AI-written content? What happens if they do?

The essay has long been an important component of the college application, enabling students to complement the more quantitative measures such as grades and standardized test scores with personal statements of interests, abilities, and passions.

Yet now, thanks to June’s landmark U.S. Supreme Court ruling on race-conscious admissions, the application essay has taken center stage.

In the court’s 6-3 decision that banned affirmative action, Chief Justice John Roberts suggested students could still use application essays to address their racial identity.

“At the same time, nothing prohibits universities from considering an applicant’s discussion of how race affected the applicant’s life, so long as that discussion is concretely tied to a quality of character or unique ability that the particular applicant can contribute to the university,” Roberts wrote.

A later clarification, however, rendered this loophole somewhat confusing. Roberts insisted that “universities may not simply establish through application essays or other means the regime we hold unlawful today.”

As such, the essay cannot become a proxy for the racial “checkbox” universities use to identify race in the strictest sense, but it can help provide an opportunity for students to address “challenges bested, skills built, or lessons learned.”

The court’s ruling, along with the somewhat ambiguous language pertaining to essays, left college officials in a fog of uncertainty with little time to prepare for the upcoming admissions season. Now that it’s here, the looming specter of illegality collides with emerging AI technology that offers exciting new opportunities but scares the bejesus out of people.

At that intersection lies the application essay.

By most accounts, the answer is no, but the technology does provide some useful results.

It’s certainly tempting to test drive the technology, feeding it essay prompts to see what it spits out. A number of folks have tried it and chronicled the results.

Writing in New York magazine, Sanibel Chai, a college admissions consultant, recounts her experience trying to determine if ChatGPT would render her services superfluous. Over several drafts, she fine-tuned her essay by prompting the bot to provide additional details and adjust the tone.

Her conclusion? It’s perfectly competent and certainly quick. But it offers “banal reflections” and “empty-sounding conclusions,” lacking originality and sounding exactly like something someone else could have written. Chai believes her job is safe for now.

Likewise, Adam Nguyen, who runs Ivy Link, a college counseling and test-prep service, found ChatGPT essays “pretty good” but also “pretty average.” The grammar and syntax are sound, perhaps a step beyond what the typical high school student might create. But it produces generic details and lacks “layers of thought.”

Nguyen says it’s fine to use ChatGPT to generate a first draft, but the hard work of editing is left to the student, especially one eyeing a highly competitive college.

“You are not going to get into a top 30 school, definitely not a top 20, with a GPT-generated essay,” Nguyen commented in the Observer.

Kevin Wong, co-founder of the tutoring service PrepMaven, reached the same conclusion.

“Admissions officers are looking for genuine emotion, careful introspection, and personal growth,” he told Business Insider. “The ChatGPT essays express insight and reflection mostly through superficial and cliched statements that anyone could write.”

Lest we assume this is the work of a lesser bot, Natasha Singer experienced similar results using HuggingChat and Google’s Bard, two ChatGPT competitors.

“High school seniors hoping to stand out may need to do wholesale rewrites of the texts they prompt AI chatbots to generate,” Singer wrote in The New York Times. “Or they could just write their own — chatbot-free — admissions essays from scratch.”

The imperative to be authentic rings especially true for students wading tentatively through the opaque waters of diversity statements.

Writing in The Atlantic, Matteo Wong contends that “high schoolers trying to navigate the nebulous admissions process may feel pressure to write as plainly as possible about how their race and experiences of racism make them better applicants.”

To satisfy admissions offices, he suggests, students will reduce their experiences to “easily understood types,” offering “tired platitudes about race.”

That’s exactly the kind of content chatbots produce, essays that will likely fail to persuade admissions officers seeking a more personal exploration of how race shaped experiences and formed character.

At the same time, chatbots lend a guiding hand often absent for low-income and underserved students. Some public schools lack sufficient college counseling services often found at private schools or publics in wealthier towns. At inner-city schools, guidance counselors are tasked with serving too many students to provide individual attention.

Certainly most such students don’t have access to the aforementioned college admissions and essay-writing tutors, who charge $300 per hour or more and, one might assume, cater to a wealthier clientele.

That’s where AI can step in and begin to bridge the gap. Singer, of The New York Times, found Khan Academy’s AI tool, Khanmigo, especially effective in helping users brainstorm and appropriately personalize and frame their essays.

“If there’s a way this tool can help those that have a different starting point catch up, or narrow those discrepancies, I think that shows a lot of promise,” Juan Espinoza, director of undergraduate admissions at Virginia Tech, told the Times.

Rick Clark, assistant vice provost and executive director of undergraduate admission at Georgia Tech, concurs.

“Here in the state of Georgia, the average counselor-to-student ratio is 300 to 1, so a lot of people aren’t getting much assistance,” Clark told The Guardian. “This is a real opportunity for students.”

To date, only a handful of universities have published guidelines regarding the use of AI tools when applying for admission. And they vary.

At Clark’s Georgia Tech, the admissions office acknowledges AI tools are “powerful and valuable” but warns students not to simply submit the raw results of their prompts.

“We believe there is a place for them in helping you generate ideas,” Georgia Tech’s website clarifies, “but your ultimate submission should be your own.”

In a similar vein, Columbia Business School permits the use of AI tools for ideation but warns students that any work submitted not “exclusively your own” results in a violation of the university’s honor code.

Among law schools, Arizona State University (ASU) now allows students to use ChatGPT on applications to its Sandra Day O’Connor College of Law, embracing the technology instead of avoiding it.

“Generative AI is a tool available to nearly everyone, regardless of their economic situation, that can help them submit a strong application when used responsibly,” ASU said in a press release.

At the same time, the University of Michigan Law School has banned the use of AI tools on applications, requiring students to certify that they have not used them. Any false statements in that regard could lead to an admission offer being revoked or even expulsion.

Proving such violations, however, is another matter.

“Will I be able to enforce it? No,” Sarah Zearfoss, Michigan’s senior assistant dean, told Reuters. “But in general, I’m relying on the honor of the people who apply in a million different ways, so this is no different.”

Savvy admissions officers might suspect chatbot-generated copy when they encounter it, but do they have the necessary tools to confirm their suspicions?

Not exactly. Admissions offices do use AI tools for a variety of tasks, including sorting applications according to grades and test scores, but filtering essays through an AI-detector isn’t yet common practice.

Could a tool such as Turnitin provide a solution? The company introduced its AI-detector last spring, allowing professors to check student submissions for bot-produced content. Thus far, the tool isn’t set up to work with admissions offices.

“We are digging around to see if any institutional admissions office has their own license, but it seems doubtful,” Jennifer Harrison, a public relations professional who represents Turnitin, told BestColleges in an email. “Educators access Turnitin via their [learning management systems]. We are not certain that an admissions office would have access to an LMS. The workflow would be completely off. Admissions essays are submitted in an entirely different way.”

For universities intent on banning the use of AI in essay writing, finding a tool to detect such submissions seems like an urgent task. No doubt enterprising tech companies will soon step into the breach and fill that void.