Essays by AI, Read by AI: Is This Where College Admissions Is Headed?

As seniors are navigating the college admissions process and writing their endless amount of supplemental essays this time of year, a new dilemma has emerged. There are concerns that application essays are increasingly being plagiarized due to student use of AI to aid them in the writing process. AI models have generated countless perfectly structured, well-written texts, which tempts high school students to use them for their college applications. Universities and application platforms have already implemented honor codes and pledges to prevent AI-generated applications. The Common Application clearly includes “the substantive content or output of an artificial intelligence platform, technology, or algorithm” in their definition of application fraud. 

One of the ethical considerations that accompany the need for these new guidelines is what our society views as the true value of the college application process. What sets apart the US college admissions process from others is grounded on the human-to-human connection, as students submit personal statements, supplemental essays, letters of recommendation, and activity lists instead of solely relying on test scores for acceptance. Most universities pride themselves on finding the most unique individuals, and those students can only show their true potential if they take the time to reflect and write meaningful essays about their experiences. Although we wish all applicants valued this process and respected its integrity, that’s unfortunately not going to be the reality since expectations are constantly rising. Students may feel pressured to use AI tools to craft competitive essays. So, should universities consider focusing almost exclusively on essay content rather than mechanics and style? Should they even create new aspects of their applications that require students to submit timed essays or video submissions?

While institutions are navigating these new challenges, they themselves are experimenting with potential ways to use AI to their advantage. These past few years, we have seen AI develop the ability to process huge datasets within seconds. The countless applications universities are receiving are the perfect tasks for these AI models. Georgia Tech, for instance, is exploring the possibility of using AI tools to notify potentially eligible students about scholarships, or to review college transcripts of transfer students, and inform them of the transfer credits they will receive in a more timely manner. Other schools, such as Virginia Tech or Stony Brook University, have started using AI essay readers which use a point system based on a rubric or AI tools to summarize student essays and letters of recommendation. This evidently helps cut down the length of the process, and they have stated they could notify students up to a month earlier than usual. 

One of the most crucial parts of the admission system is that universities look to fully understand a person, and try to incorporate a wide range of perspectives and personalities into their accepted class. During information sessions or on social media, we have heard countless times that they “aren’t looking for anything specific” as they read essays, so it is difficult to understand how they can craft a rubric for admission essays that will still allow this system to flourish. Relying on AI may undervalue the importance of this human judgement and connection, as a machine without emotion and first-hand experience with the university cannot picture how a person will fit into the school community. 

Will college admissions become more standardized again, based mainly off of test scores and school grades, and if so, what will be the repercussions? Though it may seem shocking, researchers themselves are still trying to identify how AI models take in inputs and make decisions. This causes a lack of clarity and can create issues with trust in the reliability of the admissions process, a concern that has already sparked backlash from students at UNC, for example. Additionally, the data that AI models are trained on is often biased, favoring certain demographics, schools, or backgrounds; although human readers also have implicit biases, using AI algorithms can potentially amplify these prejudices, and that is certainly something to be considered by universities. 

With many applicants using AI to write and institutions using it to read, we approach a world of AI reading AI. Ultimately, though both students and universities claim a fear of a future where AI will replace authentic human judgement, their actions are giving way to exactly that outcome. To prevent this, we must make an effort to balance the efficiency of AI usage, by implementing it as a tool to support rather than replace human decision-making in the college admissions process.

Anna Bulto '26 and Skylar Li '26