How AI Can Help or Hurt Your College Application

Webinar Transcription

2026-03-11 – How AI Can Help or Hurt Your College Application

Lonnie: Hello everyone. Welcome to CollegeAdvisor’s Webinar, “How AI Can Help or Hurt Your College Application.” To orient everyone with the webinar timing, we are first going to begin with our presentation and then we’ll have the opportunity to answer your questions during our live q and a. We have a lot of great information in store for you all.

So if you would like to follow along in case you’ve missed something or you wanna, you know, uh, refresh yourself on something that was already shared, feel free to go ahead and download the handouts. They are available. Um, all you have to do is go to the handouts tab and then click on download. So with that said, we are now gonna introduce our presenter.

Alvin: Yes. Yes. All right. So thank you all for being here today for our webinar, how AI can Help or Hurt Your College Application. My name is Alvin Carter, um, and I work for CollegeAdvisor and pretty much I’m a senior admissions consultant where I handle a caseload of students that have aspirations and dreams of going to some of the big schools.

And so I’ve had over years, over 20 years of success helping students get into college. And I’ve been on both sides of the desk, both as, uh, director of college counseling and also as an assistant and associate director of admissions. And then also working in academic affairs at Columbia School of University, well, excuse me, Columbia University School of International Public Affairs.

And so I believe this presentation is key because as we all know, um, AI is here and it is not going anywhere. So the question is no longer. If students are using ai, but how they are using it, right? Um, so for parents, counselors, and students, uh, this is a new technology that presents a genuine ethical, uh, tightrope walk, right?

So AI can be a powerful tool that can help overcome writer’s block and let’s say streamline research. But we all know the misuse can instantly undermine, um, and the authenticity and also can, um, lead to disqualification in the world of admissions. So our goal today is to kind of give you a clear, actionable roadmap for using AI as a high helpful co-pilot, not just a ghost writer.

And so we’re gonna kind of walk through the entire application process, uh, slide by slide from the initial brainstorm to the final submission. So we’ll cover what the admissions officers also known as ales are trained to spot. And so those are like the specific risks that can hurt a student’s chances.

Um, and then ethical game, changing ways AI can. Actually strengthen the application. And also I encourage you to kind of take notes as we will be diving into specific real world examples that kind of illustrates the difference between using AI as a smart editor versus using AI as a replacement for the human thought.

So it is my hope that by the end of this presentation, you’ll know exactly where the line is drawn and how to keep your unique voice. And of course, your soul is at the center of the story. Uh, the most important takeaway, I’m sorry to taking too much time, you’re, um, is that we’ll reintroduce repeatedly.

Um, okay. And what I will kind of like the college application is kind of a portrait of the human being, right? So AOs are looking for authenticity, vulnerability, and a genuine spark of intellectual curiosity. So AI can absolutely help polish. And frame your portrait. Um, but the content, the brush strokes, the colors, the feeling must be a hundred percent human.

And so if you remember that golden rule, you’ll benefit from AI while keeping your integrity and your admissions chances completely intact. So let’s go ahead and, um, dive into the poll, Lonnie.

Lonnie: Yeah, this is, I’m so excited for this topic and I, I love that metaphor that you just used, Alvin. Um, so again, we have a lot of great information as in store.

’cause I know that there’s been a lot of different perceptions and fears that have been put out there. And like you just said, Alvin, like it’s not going anywhere. Um, so you all are gonna get some great insight on how to work on your college applications, um, in an ethical and, and truthful matter. Um, so with that said, Alvin, the poll responses actually started coming in.

So we love to get a sense of what grade you all are in because it gives us the opportunity just to make sure we find ways as we’re presenting to really speak to all of our audience that’s here with us tonight. So we have 60% of our audience are in the 11th grade, followed by that. We have 23% are other, and other could be a parent, a counselor.

Another educator in some form. And then we have 6% that are in ninth grade and then 2% that are 12th grade students. So we have representation from all grade levels. So with that said, Alvin, I am going to turn it over to you to kick us off with sharing how are students using AI for their applications?

Alvin: Yes.

Yes. And so just to, you know, go straight to the polls, I’m excited to know that there’s a number of juniors or rising seniors, um, in the audience, because of course, this is a critical year and just in a few months, applications will go live and you will be writing. Um, and so hopefully this is something that you can use as a useful tool.

So we’re gonna kind of get into the first slide, which is how students use AI for the application. So let’s start with the reality on the ground, right? So when we ask students, um, we find that they’re primarily using AI in four key. Right. The first is the ideation and brainstorming phase. So to me, this is the great ethical use of it.

All right. Students often face maybe, let’s say blank page syndrome, writer’s block, and so they have a lot of experiences but don’t know what’s essay worthy. And so many of you probably haven’t wrote self-reflective essays or personal growth essays. And so this is where AI can come in handy. And so there are students that, um, I’m aware of that I work with that will feed, let’s say, large language models, LLMs.

A list of, let’s say their extracurriculars, things like, um, I was captain of the debate team or a volunteer at the local library, or taught myself coding. And then they’re asking AI to kind of generate, let’s say, 10 potential essay themes. So instead of telling the ai, write my essay, um, they may ask AI to kind of generate a list of essay angles based on my work, volunteering at the library, or focusing on themes of maybe quiet observation, um, community, community need, or kind of like the power of information.

And so this kind of gets their brain moving without writing the content for them. And I’m not sure how many of you, um, in the audience have wrote self-reflective essays, but it, it tends to be a thing where, and I’ve been in college council for years where. Seniors or, uh, rising juniors, um, gets to that point where they’re writing and it’s kind of like, I don’t know what to write about.

And it’s like, you learn so much and you’ve built so much experience and global experiences through your life, you’re not sure what to write about. And so we begin that brainstorming phase, but we all know we don’t have a, you know, everyone doesn’t have a CollegeAdvisor that, or, or guidance counselor that can sit with them all the time.

And so using AI wouldn’t hurt in that manner. And so the second area is structural and narrative outlining. So after the student kind of has chosen a theme, they kind of, uh, have a lot of scattered thoughts, which kind of often interrupt the continuity of the writing. And so they’ll input a rough paragraph about their main point or a few anecdotes, um, in kind of a conclusion idea, and then kind of ask AI, please create a three section outline that ensures my story, uh, flows logistically, um, and logically.

And so this is an effective way of using AI as a high powered organizational consultant. Um, and so I’ve had several students that kind of met that roadblock where they didn’t know what to write about. And so if they. Did an internship as a ki with a chiropractor during the summer, um, and kind of wanted to write about that, but did not kind of know how to speak about personal growth.

Um, I would kind of brainstorm with them, but on their own, just in terms of utilizing these search engines, I would also encourage them to kind of see what’s out there in terms of how to kind of outline essays and see, you know, if there’s any suggestions that you can kind of include. And so, um, a student writing about this summer job, let’s say as a lifeguard, might input the beginning, middle and let’s say the end of their story.

And then they might ask AI to suggest, um, kind of structuring it. And so the AI might structure it around kind of like the quiet routine, which is like the setup or, um, the moment of crisis, which is like the conflict or the lessons learned about responsibility, which is the resolution. And so it provides the container, but the student fills in with their personal thoughts, um, and values, or I like to save their water.

Right. All right, so let’s go to the next slide.

Okay. And the third most common use is draft refinement and polishing. And so, um, quick funny story. I noticed I was reviewing one of my students’ essay and as I’m reviewing it for the third time, I think my student forgot to erase the AI prompt that he put in there. And so when I read it, I realized that his entire statement was written in ai.

And of course he admitted it, and of course he’d done it over. Um, but I think this is where a lot of students can get lost when we speak about the draft and refinement. So, um, this kind of goes beyond the basic spell check. We are really aware of kind of like those Grammarly and things like that. So this is where students are asking AI to kind of suggest more sophisticated vocabulary or improve, um, sentence variety.

So crucially, they’re also using it to kind of tighten word counts to meet the strict applic strict application limits. Um, I think that’s probably a, a a, you know, for many students that’s gonna be approaching the application soon. You have to get used to writing compelling. 50 word statements, 150 word statements, 250 word statements, 600 word statements, and 800 word statements on questions.

And so, of course, a lot of these essays are recyclable. Um, but when colleges are asking me for 700 word essay, you may say, I don’t know how to exhaust the, the rest of what I’m saying. And so kind of, um, let’s say a student may kind of take a copy of a 700 word essay into the AI and ask, kind of rephrase this final paragraph, or to be more concise and suggest, uh, let’s say alternative verbs, um, and sentence structuring to kind of condense and kind of meet that word limit.

And so they used to kind of clean up, uh, the sentencing and theoretical maintaining their original message. So this is also, uh, the phrase where it’s easiest to kind of lose your unique voice because, which brings us to our final point of course, because research enlist building for the YSSA is important.

And so. AI acts as a high powered search engine aggregating data about specific college programs and professors or campus, uh, missions. And so this is what I love most about it. I think for many of my students. And for all of you here, you’re gonna have to kind of come up with a compelling reason to why you’re interested in a particular college.

And I know out of a lot of the top colleges, I would say a good half of those will kind of phrase a question within the supplement about Ys. And so Ys is no longer in terms of we love your school, you’re number one. I’ve always dreamed about going here. You really have to be immersed into the campus environment.

And so I feel like AI is one of those tools that can kind of aggregate this information, synthesize it, and of course give it to you in an expedite, uh, expedite process so that way you can have this information and save a lot of time. Um, so I have students kind of, let’s say a student. I had one student that, uh, was applying to MIT actually.

And so, um, I told them to kind of connect with AI in a sense and kind of find three current research projects by a particular professor that they was interested in, and also from the material science department, other faculty that’s kind of in tune. And then I also asked them to kind of search for any sustainable energy clubs that was listed.

And so the reason why I asked for this is because you’re gonna have to come up with a valid reason and authentic reason of why you like the school and the environment. So the only way to know that is if you become familiar with the school. So of course you can go to information sessions, of course you can connect with admissions reps when they’re on visit or alumni as well.

Um, but kind of nothing beats having everything right before the eye. So to see how you’re gonna angle your essay, right. So this saves a lot of time and allows you to kind of spend more time reflecting on why that research or why that club really matters to you personally.

Okay. All right. So this is, what are the risks associated with. Improper AI use on the college application. So this is essentially a warning label. So when AI moves from a tool to a ghostwriter, then the consequences can become severe. And I have seen it happen way too often, um, where the most significant risk is the loss of the authentic voice, which is so obvious, right?

Admissions officers are trained to kind of find the soul in your story. So that unique blend of humor, uh, vulnerability, and of course specific detail can only come from a human. And so AI generated text often describes, um, a homogenized or sterile kind of way of writing. And so it’s technically correct, but it kind of lacks that quirks of kind of like the student, um, memorable, um, insights, right?

So let’s say for example, an A ISA about leadership might use all the right buzzwords like initiative, uh, for certain collaboration, driving change, but it will likely omit the specific messy human detail, kind of like the student, um, that accidentally ordered a hundred. More shirts than usual for club fundraising and had to kind of hold impromptu bake sales to kind of cover the cost that messiness is kind of what ails are looking for.

So we’re looking for those anecdotal experiences that unfortunately, AI cannot produce. And when it does produce it, it is so generic, we can, we can tell, right? And so next we have the very real danger of factually incorrect information. So, um, or what we call hallucinations. And so AI models can confidently invert facts if it prompted incorrectly.

And so we, in including, let’s say, actually no, including a non-existent major, so. Let’s say, for example, a student used AI to kind of write a YS essay, um, for a state university. And the AI kind of hallucinated a program called the Honors College of Global Leadership, um, which the school probably had five years ago.

But the AI immediately knew the student hadn’t done the research, excuse me, the ao, um, knew that the student hadn’t done the research and kind of signals a lack of care and integrity. Um, so you wanna make sure that you’re up to date with a lot of information. And a lot of the time when AI is aggregating this information and kind of extracting it from different websites, a lot of these things can be outdated.

So we’re never going to lead with what’s final with what they’re recommending as far as professors, classes, student clubs. You want to use that as a baseline and then begin researching from there and then see how you can infuse your own personal thought.

The third risk is, uh, a fundamental. Fundamentally ethical one, um, ethical violations and plagiarism. And so we all are very familiar with this. Many institutions view this submission of AI generated content as a form of academic dishonesty. So the application is a promise that the work is original, um, to the applicant.

And if a college flags and essays having significant AI intervention, it can lead to immediate disqualification. And so we’re gonna get to the point in how they’re gonna figure this out because I’m sure some of the questions may be, well, how, how would they know? Um, and in, in, in some cases, you know, you’d rather be safe than be sorry, because they are incorporating and implementing new features, internal features, c through their CRMs to kind of identify some of these things if they can pick it up with the human eye.

Alright. And then finally there’s inconsistency across the application. So I’ve read maybe in my lifetime, well over 2000 applications. Um, and so when I’m looking at applications, the best part of it is the essays, right? So usually in the fall, these holidays are coming up. There’s much to do, but I’m coming home late reading thousands of essays.

I wanna stay engaged. I wanna make sure that, um, I’m reading something and enjoying the read as well. And so, too often I notice that there’s a lot of inconsistencies across the application. And so, which is a powerful human detector, right? So a polished, flowery essay that uses sophisticated vocabulary might look suspicious next to a short answer responses, activity descriptions, or standardized test scores that reflect a much simpler writing style.

So keep in mind you are self-reporting, majority of the information on your application. You’re filling in information about your activities, your honors, you have your short essays, and then you have your standard essays as well. And so when I mean continuity and flow, I mean that the syn, the flow, um, should be consistent throughout.

So even if you’re writing in 50 words, it should be the same style as if you were writing in 650 words. And so, for example, if there’s the essay that sounds like it was written by. Um, a 35-year-old novelist, but the short answers are filled with run-on sentences and basic grammar. Then the ale will doubt the, uh, validity of the entire file and one who actually wrote the main essay.

And that five shift is the serious risk.

So, uh, what do AALS care about most in evaluating applications? Uh, another slide for this. Let’s see. So in response to the rise of AI admissions officers, um, they have doubled down on what truly matters, and that, of course is human authenticity. And so they’re actively shifting their focus to look for things, uh, machines simply cannot replicate.

And even for a warning out there, I think, um, for any student that’s doing immersive high impact internship experiences, publications, things of that nature, you need to be able to speak confidently about them because these are the types of questions that they will. Be asking the interview. Um, and they’re also creating sections of the application, um, even in live videos, whether it be video essays for you to kind of elaborate.

So you’ll start to see more of that wave come along. And it’s probably because they know many students are generating information through AI and they kind of wanna connect with the students to see where is that human voice, right? Um, so first they prioritize specific lived experiences, and AI can, um, mimic general themes of perseverance, let’s say, or passion, but it cannot manufacture the visceral sensory details that prove a student has actually, um, was actually there.

And so aos are looking for the smell of the locker room, the sound of the rain on the, uh, the metal rule for the exact phrasing of a specific argument with a friend. Um, and so AI generated paragraphs might say something like, I learned teamwork on the debate team. Um. Where a human written paragraph would say, I learned teamwork huddled in the back of a bus 3:00 AM after a seven hour tournament when, uh, my partner’s voice was hos and we had to kind of rely on she bag of pretzels to kind of keep us focused on the final round.

So those little details, uh, really matter, and they keep the, the, the, the essay vibrant. Um, interesting. And it goes personal touches again, um, is what we look for. So the specificity is the proof. Um, second aos are hyper-focused on evidence of critical thinking. So they’re looking for the why and the how behind the student’s actions.

So not just the what, um, they want to see the student’s unique thought processes and how they kind of navigate through complex problems. So it’s often where generic AI summary falls short. So for example, um, instead of simply listing, I volunteered to run for the school club or school or, or excuse me, food drive.

Um, the AO wants to kind of read the student’s thought process. Like, uh, I noticed that our previous food drives only collected canned goods ignoring the need for fresh produce. I kind of spent, let’s say two weeks figuring out the logistics of coordinating donations from local charities and, and farms.

And so that process of kind of figuring out, you know, those things is the critical thinking that they want to see. Um, I’m not sure if any of you have experience in that type of illustrative writing, but again, these are humans on the other end that’s reading thousands of essays. And so we want your authentic voice and your perspective.

Okay. Third, um, AOs are looking for the student’s consistent signature style. So they need a continuous, unvarnished voice throughout the entire application. So from the essay, uh, to the activity descriptions, as I stated earlier to the counselor recommendations, they care about finding a voice that feels sincere, even if, um, it has a minor imperfections.

So those imperfections actually signal human authenticity. So we’re not looking for a perfect, well polished, uh, well, you know, capitalized essay. We’re looking for your voice. Um, and, and that’s important. So finally, the emphasis has shifted dramatically towards character and intellectual vitality. So admissions is assessing the student’s intrinsic motivation and their potential contribution to the campus community, which is very important.

You’re looking to see the things that you’re doing now. To see how you would mirror that on campus. So those experiences, um, they’re looking for, um, are those sparks of curiosity, that empathy, the personal values that deeply are rooted within you. Um, and so they don’t just want a student who got, let’s say, an A in AP history.

They want the student who kind of stayed up all night watching documentaries about every side of a conflict because they kind of had, um, to understand kind of the nuances of what it’s, that they were studying. And that’s the, you know, the vitality in ai, that, that can’t be faked. Right. So it’s important to, of course, use your own voice.

Okay. What are the biggest ways AI can help a student ethically build a stronger application? So. Now let’s look at how we can leverage this technology in a completely ethical, strategic, smart way, right? So the key is to kind of use it as a tool, not a talent replacement. So the best use is for overcoming that blank page barrier, which I call the writer’s block, right?

As we mentioned, AI can kind of serve as a powerful tool for initial brainstorming. Um, and students can kind of translate that raw experiences, their journals, their notes, their list of proud moments into a list of potential essay themes. Um, and let’s say for example, a student who doesn’t think they have a story could input their diary entries from, let’s say, a challenging summer.

Um, you know, the AI might analyze those entries and suggest the themes of perseverance, conflict resolution, and probably un unexpected camaraderie. Um, that would be the strongest. And so we want students to focus on essays where they kind of had that moment, had those moments. Um, so this uncovers compelling stories.

Um, the student, uh, might have otherwise dismissed ordinary. And let’s see, what else do we have here? Second, AI is a tremendous asset for enhancing research efficiency, so students can use AI to kind of synthesize vast amounts of information about a college’s unique offerings, such as specific research labs, faculty members, uh, community service initiatives.

Um, so instead of spending, let’s say five hours, uh, manually clicking through a university site, a student can kind of ask, you know, AI to kind of summarize three core pillars of the public policy program at Georgetown and kind of identify one faculty member focus on, let’s say, urban development. And so this kind of allows the students to kind of spend the save time to kind of truly, um, do the critical work, right?

Deeply reflecting on how they kind of fit into those programs. And then, um, kind of be the, being the human author, uh, behind the Ys, of course.

Okay.

Okay. Third, AI works brilliantly as a grammar and clarity coach. Um, I think when AI was first introduced, I think this is probably the thing that was used most with students using, um, tools to kind of check their grammar, and that was perfectly fine. Um, but of course it starts to develop into students started asking AI to kind of develop and write their stories instead, right?

So, uh, when used as a sophisticated editor, um, rather than a ghost writer, AI can absolutely help students kind of identify that passive voice, um, in repetitive, uh, kind of phrasing. So, um, or even sometimes confusing, um, sentence structuring. So, for example, a student can input a paragraph and ask, um, three sentences using passive voice and kind of suggest ways to kind of make them active.

Um, so you don’t have to change kind of the overall vocabulary or message, but this kind of ensures that the students’ original ideas are communicated as clearly and effectively as possible without corrupting, uh, their underlying voice. And finally, uh, AI is an exceptional tool for interview preparation and role play.

So students can input college’s mission, um, and their own major, um, or interest in received simulated mock interview questions. Um, so I’ve used this a lot for many of my students, and you’re gonna have, on average, I think if you’re applying to a lot of, uh, top tier schools, you’ll have probably on average maybe four to five interviews.

Um, and of course you’re gonna be interviewed by alumni. And so with alumni, they’re very familiar with the campus. They wanna see how well you’ve done your research. Um, and so you can kind of ask AI to kind of generate information, um, on questions that could be potentially asked. And of course there’s a lot of websites on Google, but I think if you know how to use, you know, prompt it, right?

Um, and you’ve kind of given your details and things that you’ve done and things that you’re proud of, you can kind of, um, see, you know, what questions would be your strengths and what questions you should work on, um, from that. And so I know AI can kind of simulate, let’s say, a mock conversation so it can ask you tough questions like, tell me about a time you failed, and what did you learn or how will, um, your specific passion for music kind of impact our biology program?

So this kind of allows the student to kind of practice articulating their thoughts, um, out loud and kind of building confidence before the real interview.

Okay. Uh, where is the line between helpful tool and misrepresentation in an application? So this is where we kind of get a little serious, right? This, this slide is kind of the heart of the ethical dilemmas, kind of like where exactly is the line drawn and kind of how we define the boundary by four core concepts.

The first isal control, right? The line is crossed when the AI stops being a tool and starts being the author. And so a helpful tool supports the student’s own voice, but mis misrepresentation of course, replaces it with external one. So let’s say for example, if the student cannot explain the reasoning behind a specific sentence or why they use a certain literary literary device, um, it’s a huge red flag.

So if an AO acts what inspired that specific metaphor and the student had to say, well, you know, the AI suggested it and the student has kind of lost that author. Uh, authorial control. All right. The second boundary is between idea generation and versus content creation. So using AI to brainstorm a list of topics is generally considered a helpful tool.

Um, but allowing AI to write the actual narrative, the anecdotes, the failings, or the specific dialogue is misrepresentation. So the meat of the essay, the feelings, the specific sensory details, uh, let’s say the personal growth must come directly from the student’s own memory and reflection. And so asking j uh, AI to kind of give you three ways to kind of start an essay about community service is fine, right?

But asking ai write the first paragraph of an essay about the time I volunteered at a soup kitchen is a violation.

Third is unearned sophistication gap. And if a student uses AI to kind of, uh, inject highly advanced vocabulary or complex philosophical concept. They do not actually understand or use in real life, um, it kind of constitutes a form of misrepresentation. So I wanna make this very clear, um, what we’re looking at when we’re re-read the personal statement.

Of course, it’s almost like a writing sample. Um, so we need to see how well you write. Of course, we’re looking for your insights and your perspective, and then we’re looking to see if you’re a perfect fit for the institution. Um, so you don’t wanna kind of cross those lines of kind of losing that voice at all.

Um, and so I wanna make sure that we all know that misrepresentation can mean a lot. We’re gonna go into it in a few, but misrepresentation can mean a lot. Um, and we’re starting to see a lot of students get rejected and it’s hard to call. It’s not like you can call the admissions office and say, why did I get rejected?

But from my insider scope, um, there are universities that are chopping down on it because they’re realizing that, um, it is happening too much. So, uh, this kind of creates a false, misleading image of the. Academic level. Um, so for example, a student who normally writes, let’s say at a 10th grade level, might have an essay riddled with words like, uh, synergistic or ostentatious, um, or juxtaposition.

And so we’re not looking for nothing for, we’re looking for your voice. And so, yes, you wanna be sophisticated in a sense, um, but you wanna be careful with that as well, because even if you wrote it on your own, those are red flags that will have, uh, application pulled just to review to see if it’s been, um, generated through AI at all.

So this false sophistication is easily spotted by an ao and it’s a form of dishonesty about the student’s true capabilities. Um, so just be yourself. Finally. The line often involves transparency in institutional guidelines, right? Um. Crossing the line frequently kind of involves like ignoring the specific AI policy set by each individual college.

So if the school asks for a statement on AI use and the student kind of hides or minimizes the extent of the tools involvement, the lack of transparency kind of moves the usage from helpful to dishonest. Um, so just know that everything you’re doing for college, when you’re signing your signature, read the fine print, it says everything there that you are to be honest, and everything is to be true and that you’ve read this four or five times over to ensure that.

Right.

Okay. So let’s see. What do colleges AI policies typically say and what families assume if the school’s policy is unclear? Alright, so college AI policies are patchwork right now. Um, so it’s really hard. To be clear, and I, I highly suggest, you know, when you connect with your regional representatives, if you are in high school or counselors, if you can connect with, uh, admissions officers at set schools to kind of see their specific guidelines, um, I would highly recommend doing that because there’s some schools that’s a okay with using AI for some things.

If there’s some schools that’s totally against it. So we

see a spectrum of policy, oh, AI use the kind of guiding that integration, which AI can be allowed for brainstorming or editing, um, within that disclosure. So most top tier schools fall into middle ground, um, encouraging students to use AI as a tutor or coach, um, while insisting the final product be the student’s original work.

So the key is that most policies value the student’s intellectual honesty and that their originality of the thought. Alright. Um, what slide are we at now? Let’s see. Okay. And this kind of leads to the critical element, right? So the disclosure requirement, many institutions now kind of include a checkbox, a short response prompt, asking students to kind of describe if and how they used AI in the application process and falling or failing to disclose ai.

Um, when, when prompted is often viewed as more than serious offense, um, that can actually violate the honor code and you can be rejected. Now, I haven’t heard of any stories of where universities communicated this to anything else, but if a university finds out that you used ai, they could report this to the common app or whatever, um, portal that you did use, I’m not sure what the extent of that would go, but I would hate to risk by chances of attending all these schools just because I didn’t double check something.

So very, very, very important.

Okay, let’s see what we have here. All right, so. What should you do if the policy is vague or non-existent? Right? And so we strongly recommend the assume minimal use rule. And so if a college’s policy is unclear, the safest and most strategic assumption is to treat AI as a severely restricted tool. So families should assume that the institution expects a hundred percent human authored essays, and that any AI involvement should be limited strictly to the high level brainstorming, structural, outlining, or let’s say basic spell checking.

It’s always better to err on the side of the caution, um, within ity as well. Lastly, um, and this is the most critical point, is the honor code implication. So regardless of specific AI wording, almost every college application submitted under general honor code or statement of authenticity. So when you sign that application, you are legally and ethically attesting that the work is your own.

And so improper AI use where the machine. Is the author is a direct violation of that oath. And so, for example, if a student, let’s say, is accepted, uh, ’cause I get this often too, like what, what if I’m accepted? Uh, if you’re accepted and the school later discovers the bulk of your essay, has AI written or was AI written, then they have the right to kind of rescind that offer.

Um, and because the student violated the honor code they signed, um, that that’s grounds already. So wherever you sign your name, um, you’re holding yourself accountable to.

Alright, so do admissions offices use AI detection software and nip? So how reliable is it? Right. This is the question on everyone’s mind and the answer is very complex. So regarding current use patterns, um, while some offices are kind of experimenting with AI detectors like GPTZ, um, zero or turn it in, um, many elite institutions are, are hesitant to kind of rely as they kind of primary.

You know, primarily using the human centric part of things, right? And so they’re more likely to use them as a flagging mechanism to prompt, um, to a second more critical human review of the application. So what that means is, let’s say I’m re reviewing application and I realize something is not right, and there’s something that’s red flagged, I may say, you know what?

I, I wanna pull this application and I wanna put it through the AI services that we have. Right? And so a lot of them are internal, a lot of them are building in strong programs as well. Um, so, you know, I hate to scare you guys, but you wanna be on the safe side, the sorry side. So, um, the main reason kind of for this hesitation is kind of the, uh, reliability gap.

And so AI detection software is notoriously prone to kind of do false positives. So the reason why they don’t really have anything effective now is because, you know, you would have so many students essays being pulled and red flagged, um, because there’s certain things that AI picks up on that could be normal, that for some people it just, you know.

They, they voted. So let’s say it often kind of flags the writing of non-native English speakers. So highly formal writers are those who kind of use, um, very common structured phrasing as AI generated. So because of this inaccuracy, most admissions directors, um, view detective results as a circumstantial evidence at best, rather than a definitive proof of cheating.

Um, so for example, a high schooler who’s, let’s say main language, um, is in English, who writes in a simple structured subject verb object style, um, is statistically more likely to be falsely flagged than a native English speaker who uses, let’s say, complex syntax, even if they use, even if they both use ai.

Okay.

Okay. In fact, the, uh, human detector is more effective. Um, experienced admissions officers are often better at detecting AI than the software is. And so they notice the lack of genuine voice, the repetitive machine, like sentence structures, and then of course, the absence of specific perky human details.

Then the vibe check by a seasoned reader, um, is the one who will take a read, and then they’ll just see if this, if it, if it’s right. And so this is just about the feel, right? So it remains kind of the most powerful tool in the admissions office. AR school is the human right. And so, for example, a human AO will immediately notice if a high school students essay repeatedly uses the phrase in conclusion or, um, is imperative that which are common AI artifacts, but are really used by teenagers in personal essays.

So be very careful too of your style of writing. Um. As well. And then finally, the shift is toward holistic verification. So instead of relying solely on detectors, officers are cross-referencing essays with other parts of the application, such as, um, let’s say grade school papers. I’m not sure if you know Princeton University, um, requires a graded paper that you had either from the past three years or where while he was in high school.

Um, and so I’ve read some papers where students wrote pretty well. Um, but the, the, the level of writing was so clear of a difference, um, in so much of areas of improvement that I told them, you know what, choose a paper that kind of reflects your writing now. Right? You don’t wanna choose something from ninth or 10th grade because you may have grown.

Um. In terms of your writing ability. Um, so you wanna be careful with that as well. So they’re looking for a consistent fingerprint of the student’s writing style that should remain consistent across all documents. So if the essay is a work of Shakespeare, but the teacher recommendation describes the student as a developing writer, then we have a serious mismatch.

Mismatch. And, and teachers do write like that because we want teachers to kind of speak of how you are academically as a critical thinker, as a writer, as a public speaker. Um, and so if it comes across that they haven’t mentioned that you are, um, a Shakespeare writer, then I think it’s time for you to kind of, um, windle down and, and brainstorm a little bit and try to make sure the writing is more fitting to your age, your grade level, and to your voice.

Okay. How can students preserve their authentic voice if they’re using ai? SAA, excuse me. I say brainstorming, outlining, and or editing. So this is the how to slide for students who want to be smart about using AI without sacrificing, um, the application’s most valuable asset, which is your voice. And so this first technique is the vulnerability in injection.

So students must ensure that they include, let’s say, moments of doubt, humor, or specific sensory details that an AI cannot generate. So adding a sentence about a specific inside joke with a sibling, or the exact feeling or failure after a missed goal kind of provides that human anchor, um, that grounds the essay in reality.

So for example, um, if AI suggests a formal sentence like, um, I demonstrated resilience in the face of academic hardship, the student could change it to something like, I failed the physics test so badly. I almost threw my backpack in the dumpster. Um, but I went to kind of tutoring center, um, and kind of worked through that, um, stubborn ways and kind of, you know, inch on to kind of.

Propel, then yeah. Uh, we would want to see that storyline better than just a fancy sentence. That kind of concludes those things. Uh, second, we want you to kind of employ, uh, iterative prompting and personalization. So instead of starting with the blank state and asking AI to kind of write an essay and state asking, um, to kind of like write an essay about soccer, or students should kind of provide essays in their own real voice.

So never kind of go to AI for the use of kind of just like, instead of brainstorming, you’re just asking them to write your essay because the student, um, could kind of input let’s say a two page stream of consciousness, journal entry about a fight that they had with a friend. And then ask AI to kind of identify the lesson learned.

That’s good. But when you are technically prompting and telling AI exactly how you want the essay, um, exactly how you want your, your, your style, um, your syn takes your flow to be, um, then it becomes dangerous. So this doesn’t sound like you, then it’s gonna be a red flag. Finally, students should kind of, um, be comfortable maintaining micro flaws.

I think this is something that, um, adds a, a great value to the humanistic side of things. Like a perfectly polished essay, uh, can often feel suspicious. So students should be willing to kind of keep some of their unique writing quirks. Um, there are signature style phrasings or even like metaphors that’s kind of slightly less than perfect.

Um, these micro floors are actually badges of authenticity and they kind of signal to the reader that the work is crafted by developing, um, a developing young writer. And so, um, you know, keeping a slightly mixed metaphor, like, you know, that experience open a door to a new path is better than letting the AI turn into, um, a sterile, that experience provided an infection point, uh, of future development.

So, of course, be careful with the wording.

Okay, how should students use or avoid AI for non essay sections? Um, so the essay is the focus, um, but the best and the rest of the application matters. Um, best parts of the application too. So AI can kind of help use for many different reasons, right? We don’t only have to look at the, um, essays. You can look at the activity descriptions, which becomes really popular too to kind of help students with action oriented, um, extracurriculars.

So this means kind of like turning passive phrases into strong impact, uh, focus bullets. So for example, a student can kind of input, uh, I was the leader of a French club, and kind of ask AI to kind of generate a more action oriented bullet point. Like kind of like spearheaded three cross-cultural exchange events.

Um, increasing club attendance by, let’s say 40% over one semester. However, students must be careful not to exaggerate their roles or buzzwords that don’t accurately reflect their actual level of involvement. And so that’s what we call in the work world, embellishing or, you know, exaggerating, um, on let’s say your resume.

And so you wanna make sure that you’re using words that you can really speak to, right? Um, so in contrast, the additional information section should also remain strictly human author. So I know I said my favorite part of reading applications are the essays, but I think the one part that I enjoy the most is probably the additional information section because, um, it’s a part of, aside from everything else that’s so, um.

You know, cut and dry. Uh, here we have a section where it’s kind of like, if there’s anything additional that you’d like to share with us, please do. And so this section is often where students kind of explain personal hardships, medical issues, or kind of unique circumstances. And so using AI here can make a deeply personal explanation feel cold or disingenuous.

So potentially kind of undermining the empathy and a might over, um, otherwise, uh, otherwise fail. So for example, um, an AI written explanation of a family emergency will often use the sterile, formal language that kind of prevents the human connection, where a student kind of can, you know, make that sound much better by using their own.

Okay, so students also, um, should also absolutely avoid AI for email correspondence, um, with AOS or department heads. So I know that we do have some freshman, sophomores, and juniors in here. I know that you’re not facing the application process yet, but this even starts earlier than that, right? All of your correspondence with admissions reps or faculty should be your own voice and not ai.

Of course, you can have a template and kind of work with that, but these interactions are opportunities to kind of build a personal rapport. So a templated, uh, AI email is easily spotted and, um, and suggests kind of the student’s not truly interested enough to kind of write a three sentence message themselves.

And, um, that’s important. Finally, uh. The interview, thank you notes. Uh, you’re gonna have interviews for several things, whether it’s for an internship, whether it’s for a summer program for some, right? And whether it’s for colleges. And so you’ll often have the time, have time to kind of connect on a personal level with someone.

And your writing does matter, right? We still want it to be, um, impactful. We still want it to be compelling, but we don’t want it to be flowery. And so thank you. Notes must be personal and kind of reference specific moments from the conversation. So an AI generated thank you notice often generic. Um, thank you for your time.

And it kind of fills for the mention of kind of like the unique sparks of the interview, such as kind of a sheer joke or specific question about a professor’s research. So this makes the student kind of appear transactional rather than genuinely appreciative. So, um, my years at Columbia University, I’ve had several thank you emails.

Um, and of course many of them were of course just the regular, same off, same out. But nothing beats a handwritten letter. Nothing beats a letter that is just written in the student’s voice, um, very casually. Um, just to connect, uh, because it allows the other person to kind of let their walls down and kind of, um, you know, let their guards down rather, and kind of feel comfortable and release or put you at ease in those 10 situations.

Okay, let’s see what we have here.

All right, so I think we are close to finishing and so we covered a lot of ground today and I want to kind of leave you with four final pieces of strategic advice. And so first, focus on self-reflection First, right. Before even opening an AI tool, students should engage an analog brainstorm. Click, get a pen, get a paper, identify some core values, major life lessons, most compelling stories.

Um, the strongest applications are built on the foundation of deep independent self-reflection that no technology can replace. Um, so we want you to kind of, um, let say, for example, access student, what is the one story you would tell a complete stranger to help them understand who you are and write down that answer.

Um, I currently have all of my juniors now, um, well, they’re already right at the beginning of this year. Um, why this major, why this school? Just to kind of get them to exercise their brain a little bit on answering these types of questions. Because you have to think of it this way. You’re applying to, let’s say, six or seven schools that may have a similar question.

And so you may have a y essay for this question, and you’re like. Okay, how, how, how do I write that? I continue to write this, the other ones, and we can use that same formula approach, um, to it, right? Second is embracing the human in the loop requirement. So always ensure a human is involved, right? A parent, a counselor, a teacher that’s gonna be able to review the final application and provide a reality check on the voice and tone.

So a second peer humanizes the best defense against ai. Um, and where the machines kind of influence slowly, kind of takes over the student’s original intent. The editor, um, can kind of, or the person that’s kind of given feedback, um, can kind of tell ’em, Hey, you know, I, I can’t hear your voice in, in this statement.

So if you do feel that way, ’cause I do have a lot of students that, that do push back as well. Like, I’ll tell ’em that you need to work on something. I’m like, no, but everyone said this is good. Um, understand that you need to take critical, um, suggestions as well, right? Um, critiques. And so be open to that human side of things.

So even if you’re developing something and you use AI to kind of even outline. You still kind of wanna connect with, um, someone, primarily a counselor or a teacher. Um, but the ones who know you the best can really tell if it’s your voice. So I think it’s time to knock on mom and dad’s door and say, Hey, can I read you, uh, my essay?

Although I read several essays with students, don’t want the parents to read. Um, you kind of want to see who knows your authentic voice. Uh, third, uh, remember that the goal is quality over efficiency. So while AI can kind of speed up the process, um, the goal of college applications is not speed at all, right?

Um, but connection. So we want students to be willing to kind of take on that slow path for the most important parts of the application, ensuring that every word serves as the purpose of showing who they are and, um, what they value. So, uh, spending an extra five hours on a single 650 word essay that is a hundred percent authentic is always gonna be better investment than generating a five separate.

AI draft in an hour. Uh, finally, our own your own, own your own narrative. Um, ultimately the application is the student’s first adult responsibility and, uh, a chance to kind of tell their own story to kind of the world. And so using AI as a helpful consultant is a modern smart. And so owning the final product with pride and integrity is what, what truly went over in the admissions committee.

And so the person they admit is the person who wrote the application. Make sure that the person is your student. Um, and so we wanna make sure that you put your best foot forward. Thank you all. I’m happy to kind of take questions now. Um, gotta a little tongue tied there a little bit because I forgot to drink some water, but I’m happy to take questions.

Um,

Lonnie: yes. Yes. Alvin, this was excellent. So I am gonna give you a water break ’cause you have shared a lot of great information with our audience. And so I’ll go over just a, a couple of reminders before we jump into our live question and answer. Uh, so, um, my reminder is that you can download the handouts in case you missed anything.

Um, feel free to download those handouts. And I did get a few questions in regards to is this webinar being recorded? And yes it is. And so it will be available to you in your email. So just give us a day or two and you’ll be able to access the link to go back and watch the recording. All right, so now we’re moving into our live question and answer.

Um, we may not get to every question. There are a few questions that have come in. I’m gonna definitely, um, you know, put them together so that Alvin can be able to answer them. Um, the first Alvin, are you ready for the first question? Sure. Hey, great. So I got a few questions in regards to, um, what happens when.

Uh, with international students writings and many opportunities, AI detectors tend to say the writing is done by an ai.

Alvin: Yes, yes, that’s true. Um, and the very reason why admissions offices are really not locking down on this whole. Elimination of, um, students that may be pulled up from the detection of ai.

Um, understand that in different countries there’s certain nuances and even in words, um, verbiages and things like that, that can be different. And so that may often pick up, right? And so I do wanna make this clear, right? There’s a level of what will get you disqualified. And so if you’re using ai, I’m sure that there’s a skill that they’re using, whether it’s kind of like, well, this person clearly used AI for the entire essay, and things may pull up where you may have used it for Grammarly or other things, because there are detections that can pull up on every little thing you do.

So I think the goal here is not to find a reason to disqualify you. It’s just to make sure that if you are using it, you’re using it, right? Um, and so if you are attending a foreign school in, uh, out of the country, and the American language is, well, not language, but in terms of the writing style could be different, that’s where the human, um, power comes in, right?

See if you can have a counselor to proofread or you can use AI to say, Hey, I attend school in such and such country, um, and I’m up writing for colleges in the u uh, us. Please make sure that my tenses in language is, you know, you know, compatible or relevant. Um, or the tenses is a relevant rather, so that shouldn’t be a problem at all.

Um, but double check. Always have someone look at it.

Lonnie: Okay. Thank you. Um, so someone, uh, said, thank you so much for all this wonderful information. Thank you. Uh, can you repeat what you stated on the onset at the beginning of the presentation, you had given the metaphor around like the art, you mentioned something like effective AI versus humans as colors, uh, brush stroke.

Do you remember that metaphor?

Alvin: No, but there’s a good thing. The great thing is that you’re gonna have the recording. Um, and I’m gonna have to look at it too because, uh, I was looking at a tiny screen and I was just trying to present. So I, there’s a lot of things I did forget. Um, but yeah, you’ll, you’ll definitely have this video to kind of look back on, but, um, if there’s anything in particular, uh, I would love to answer it, um, about Sure.

Lonnie: Absolutely. No, and I, I’m getting a lot of feedback around just. How useful this presentation was. Someone just shared it. They learned a lot. Um, so definitely Alan shared a phenomenal information. So you will have access to that recording and you could go back and you can catch that, that metaphor, uh, because I, I don’t remember exactly either, but it, it definitely resonated with me too.

Um, so the next question is, do institutions use AI checkers when going through the essays? It’s very possible that when a student runs their essay, um, that they’ve written all on their own through an AI checker, it shows that 30% or even 80% of it is AI generated. Sometimes there’s only certain ways to express and experience even though it’s uniquely individualized.

Alvin: Great question. And you know what’s interesting is, um. I’m a writer myself. And, um, I, I look at people’s writing and I have my own systems to kind of check my students are are doing anything. And there was one student I had that it kept popping up that they used ai and she literally cried on the phone with me and said, Carter, I swear to you I did not use it.

And so I did an investigation and realize that that does happen often where there’s, and that’s why we’ve been open in the presentation, there’s certain phrases and things that, um, are generic, um, to ai that AI may pull up. And so that’s why I kind of counted it with, you know, the schools does not, does not have a hard policy on it.

Um, and so let’s say for example, 30% of, if in a world of where they do have metrics for this, let’s say 30% of your essay was probably AI generated, um, I don’t think they would disqualify you for that, right? Um, depending on how it, it is written, right? Because they do understand that there’s certain languages, there’s certain phrases, um, that’s gonna, that’s gonna pop up.

So to be on the safe side of things, I would say no. But to answer the question, just straight to the point, universities are not literally just pulling essays. Um, they’re only pulling it if it’s red flagged, right? So if I’m personally reading an essay and I’m realizing that, hey, this person wrote their descriptions for the application, but it, the style of writing does not match to this beautifully written supplemental essay on Ys, I may say, you know what?

Let me pull this. And so the thing here is that it’s very easy to do. The only thing you have on your side is that we have an influx of applications that we don’t wanna pull every application because that means we’re gonna slow down our process. But you do not want someone to look at your application and say, you know what?

Because if they are pulling it, they’re gonna find something. ’cause they’re not gonna waste their time pulling an application, um, just to put it back in the minute and say, oh, nope, that was 20%. So, um, there’s more nuances to it. My best, uh, answer to give you though is to connect with the universities that you like to go to and explicitly ask those, those questions.

How is a I used in the terms, uh, in form of admissions here? Um, and to what extent?

Lonnie: Okay. All right. So I’m gonna take a short pause, um, for my question and answer, so feel free to continue to ask questions. I wanna share more about the work that we do here within CollegeAdvisor. For those in the room who aren’t already working with us, we know how overwhelming the admission process can be, and even just the stress around AI usage, what’s appropriate, what’s not appropriate.

CollegeAdvisors team of over 300 former admission officers and admission experts are ready to help you and your family navigate the college admission process in one-on-one advising sessions. And essay editing Through our digital platform, CollegeAdvisor has had 10,000 total lifetime clients and a 4.8 out of five rating on Trustpilot with over 750 reviews.

After analyzing our 2023 through 2025 data, we found that clients working with CollegeAdvisor who have a strong academic and testing record are 2.6 times more likely to get into an Ivy League around three times more likely to get into a top college and around five times more likely to get into John Hopkins, UCLA and NYU when compared to national acceptance rates.

Increase your odds and take the next step in your college admission journey by signing up for a free 60 minute strategy session with an admission specialist on our team using the QR code on your screen. During this meeting, you’ll receive a plenary assessment of your academic profile, along with some initial recommendations on what you can do to stand out.

At the end. You’ll also learn more about the premium packages we offer that pair you with an expert who can support you in building your college list, editing your essays, and much more. We are here for the whole process. Okay, so with that said, we’re gonna take a couple of more questions. Um, so let’s see.

Someone’s asked, how do you resist a temptation of using AI to write your essay?

Alvin: Hmm. Uh, let’s see. Well, I wouldn’t say resist, right? I, I, I wouldn’t, you know, experiment with it. I would and, and, and that’s why I opened it up with, you know, and AI is here to stay. Um, and so. We’re looking for critical thinkers all the time.

So you have to maintain that level of, um, authenticity when you’re, when you’re writing, but it’s okay to use it for the following reasons that I said to, to brainstorm ideas, to help outline your essay. My favorite thing I use it for, and I find it so effective, is the why us, why School essay, because you can literally put in all the things that you’re into and kind of see, navigate how you can go about doing that at the school.

So when you’re writing your YSA, you’re connecting it to just student life, the academic experiences, the civic engagements, and you can kind of have like a one-stop shop in that. But as far as writing statements, you know, this is just me personally, I really feel like brainstorming human to human. Nothing beats that.

Nothing beats that. Um, and if you find trouble doing that and you wanna experience with ai, AI is a growing tool. And so I don’t want you to be laggard behind in it in terms of, oh, I know we can use that for that. Um, but again, don’t. Putting anything about help me write this sentence or help me come up with this paragraph.

Or I have this story that I wanna write. Can you put this together for me? Just little, little things, right? And adults use it too, right? We wanna make sure that when we send things that the grammar’s correct, we wanna make sure that we have a compelling statement. But using it for the sole purpose of having them generate any sentencing for you and then you using that, um, is not good.

I think the only time, the only wiggle room I would give a student if I’m reviewing the application is probably in the activity list section, right? Because here you have, um, things that you’ve done since sixth grade, since ninth grade that you have to synthesize in 150 characters. And here you have all these amazing things and you’re like, Carter, how do I put this to a literally 150 characters?

I didn’t say words instead, 150 characters, which I’m sure many of you are aware of. And that becomes complex. So using AI for that would be amazing, right? Um, not to say ai, can you shorten this for me or make it 50 words, but. What things can I get rid of? How to synthesize, how to quantify this information is perfect because when I’m reading an application, I, I just wanna see your impact and is it quantifiable, right?

If you’ve done something, how, what’s the context of it? How did you impact others, et cetera.

Lonnie: Okay. Uh, this is gonna be our last question, and it says it’s very difficult to find the right place between voice and what an admission counselor might see as low writing skills. How do we determine what to keep or removed?

I bounce my writing off my mom and I find her recommendations removing what I feel is my own expressive style.

Alvin: Ooh. You know, that’s, you know why? Here’s the thing, get used to it, right? Because you’re gonna have colleges where you have to literally write a, a full, like you’re gonna have to write a, a statement in 150 words.

Um, or even sometimes even for Columbia University, I think there’s four statements are like, no more than 150 words each. So I think, um, balancing ideas off of each other is great. Um, I think using AI to kind of help with some of those questions are, are great. But, um, and what was the question exactly?

Again, I wanna make sure I, I answered the tail end.

Lonnie: Like, how do you, how do you kind of make it, you have your own voice versus your mom telling you, say it this way, say it that way, but then the student feels like

Alvin: it’s ah, ah, that’s, you know, that’s what I was gonna say. I was gonna say, you know what, that the reason why I was gonna say you have 50 words to have a lot of students that, um, once they go through the services, even with us, because we have a whole milestone team.

To help students with their writing is that once they get their stuff back from the essay review team, they’re like, but that’s the favorite part of the essay that I don’t wanna remove. And they don’t understand that you have to be concise and get to the point. So the only time I would ever tell a student take something out, even if it sounds beautiful, beautiful, is that it has to be concise.

And another thing I wanna leave you all with is that you have about maybe eight, nine minutes tops to get your application reviewed. And so the advice we give you of being concise and cutting things out and getting to the point is strategic.

Lonnie: Got you. Got you. Okay. And then they, they kind of said Mom and AI are the same.

Uh, um, that was like a, a clarifying point in their question, but thank you so much audience for your engagement. Thank you for your feedback. I know there was a couple of, uh, comments I about, well, you didn’t get to the last slide. I really wanted that third point, um, that Carter was gonna make. Um, and again, we do have the slides that are available for you, so please go ahead and download them now as I am closing out the webinar.

Thank you everyone for attending. Thank you Alvin, for this presentation. I feel like we need to do a part two of this because it, I agree. It was, it was really, really helpful. Um, so with that, everyone, when you close out your page, you will have another screen that will pop up. So if you have not, you know, signed up yet for a free assessment with us, please keep the browser open When the webinar ends, you will have the opportunity to sign up for a free assessment with us.

Alright, everyone, have a great one. Goodnight. Bye.