Letters of recommendation (LORs) are a teacher’s description of a student, their passions, their goals, their presence in the classroom–everything human. Artificial intelligence– the nonhuman. Now, the two have become intertwined with the introduction of AI onto Naviance, a website geared towards informing and preparing students for their futures in college. 

The new AI elements include PowerBuddy, an AI assistant; improved College Search and Career Exploration; Teacher AI and Counselor LOR AI. Teacher AI and Counselor LOR AI are meant to facilitate the composition and drafting of LORs for staff. All of these AI features can be enabled or disabled by the district.

Oak Park High School’s administration disabled these features for staff this past fall. However, PowerBuddy features for students remain active.

“When Naviance introduced an AI drafting feature for teacher recommendations in summer 2025, it was rolled out without prior notice to the school or an option to opt in or out,” Amanda Fitts, OPHS college and career counselor, wrote to The Talon. “Once staff became aware, the school reviewed how the feature functioned and how it could affect students. Because the tool could generate draft recommendation letters using limited profile information, the school chose to restrict it to avoid confusion and protect the integrity of the recommendation process.”

Whether AI should play a bigger role the college admissions process or be eliminated altogether is up to debate, so here are some of its pros and cons.

Pros:

Efficiency is one of the primary benefits of AI. According to a UTHealth Houston, AI models can “process syntax and semantics at a speed and scale that far outpaces human capabilities.” The admissions process is full of tight deadlines, and it would seem more convenient to turn to AI for support.

This is the exact purpose Naviance has intended for their AI usage. The new features are meant to alleviate the stress staff may feel when faced with the extensive profiles of their students, and having to pick through all the information to write a cohesive letter. The AI can “[draw] from each student’s full profile and writ[e] in a voice that stays true to the author’s style.”

Of course, it is to be noted that AI merely provides a first draft; the teachers and counselors have free rein to edit the generated letter and add their own words. In some cases, AI is used to brainstorm, but the actual writing is done by humans. 

AI-generated letters can also eliminate bias in LORs. In a paper written by Fried, Johnson, Pelletier, Landman and Bartz, it is established that LORs written for women and racial minority applicants skew towards personal characteristics rather than achievements. AI can standardize LORs, either by providing the initial framework for the letter or by checking for any bias after the letter has been written by a human.

Cons:

Despite these benefits, the use of AI in letters raises the typical concerns. AI is not human and as of yet does not contain the same emotional depth as a human author. Thus, an AI-generated letter may not truly express a teacher or counselor’s thoughts on their students, therefore impersonalizing something intrinsically personal.

UTHealth Houston goes on to state that AI “lacks genuine creativity, intent, and emotional nuance” compared to humans. A strong LOR should incorporate personal anecdotes and experiences the writer has had with the student. If AI is unable to capture this, it can be argued that AI-generated letters will result in weaker chances of admission.

“Recommendation letters are meant to provide colleges with a human perspective, and OPHS teachers and counselors spend a significant amount of time personalizing each letter to help admission officers understand each student’s unique strengths and attributes that will benefit a college community,” Fitts wrote.

Fried, Johnson, Pelletier, Landman and Bartz’s study states that LORs remain one of the “few robust, holistic metrics for in-depth insight into the character and quality of the applicant,” especially with the prominence of standardized testing and black-and-white grading. Whether AI should have any correlation with such writing is up to the individual and to the school district. But regardless, LORs should be authentic reflections of a student, involving the collaboration of its writer and subject.

“Students should request recommendations from teachers who know them well and whom they trust to provide an honest, personal perspective,” Fitts wrote. “In some cases, this may mean a student asks a teacher from a class in which they experienced difficulty. If the student consistently attended Support Time and demonstrated growth, the teacher can provide colleges with valuable insight into the student’s effort and approach to learning.”