Can Law Schools Tell If AI Wrote My Personal Statement?

I wrote the essay myself. I only asked ChatGPT to make it sound better. Could that get me rejected? Or maybe you haven't used AI at all, but you ran your personal statement through a detector, and now it's telling you that it looks like you did use AI. So can law schools actually tell? Sometimes readers may suspect AI use, but a detector's verdict isn't proof. And the risk I see most often is actually harder to spot. You think AI improved your essay because the writing sounds better, But the substance and the details that actually could have made you stand out are now gone. I'm Mara Frag, an Ivy League-trained lawyer and the founder of Top Law School Consulting, one of the top-rated law school consulting companies in the country. I can often recognize when a client has just put an outline I gave them into ChatGPT. The structure is there, the experiences are theirs, but the places where we needed their most interesting thinking have been filled with overly polished language and it often tells me very little. It's surface level. That's what I want you to understand before you submit. We'll cover detection and what counts as assistance technically from law schools, but also most importantly how to recognize when your essay just sounds more impressive but has actually become a weaker application. Okay, first let's talk about can schools actually tell if you've used AI? There are three different questions here. Can someone prove that AI wrote your essay? Does your essay read like something generated by AI? And third, does the school allow the kind of assistance that you used on your essay? Those questions have different answers. AI detection tools, they make mistakes all the time. We all know that. So a detector score is not definitive proof of who actually wrote an essay. And a reader's suspicion isn't proof either. So I wouldn't tell you that every school can reliably identify every AI-written essay. I also am not going to promise you that no one's going to notice if you use it. Experienced readers recognize patterns. If you spend enough time reading these essays, certain rhythms or structures, there's just ways of describing experiences that start to become familiar across multiple applications. But if you wrote your statement yourself and a detector flags That result alone definitely doesn't establish that you did anything wrong. So don't panic. Don't assume you need to make your writing all of a sudden worse or strip out every em dash to prove you're human. My clients panic all the time about this. Keep your notes and drafts. Focus on producing a thoughtful statement in your own voice rather than just repeatedly rewriting it to satisfy different detectors. And of course, check each school's current instructions. Schools have different rules about AI and outside assistance, and they're constantly changing right now as AI is becoming more prevalent. So don't assume something's permitted because another applicant did it or another school is okay with it, or because you only used AI for part of the process. Whether assistance is allowed and whether it improves your application, though, those are separate questions and very important ones to think about. What makes an essay sound like AI? When I'm reading a draft it's usually a combination of things. It's the language is just a little overly polished and also strangely impersonal at the same time. The paragraphs often follow a predictable rhythm, actually even sentence structure often has a certain rhythm that AI loves to use. Also every experience leads to like this perfectly packaged lesson. Something difficult happened. It taught the applicant, let's just call it resilience. They helped someone and it showed them the importance of advocacy or they encountered an unfair outcome and it inspired them to pursue justice. These are too neat. The ideas might be sincere, but they're also conclusions that thousands of applicants could reach. So what's missing is the thinking, that critical analysis that makes your experience specific to you. What did you initially understand? What did you initially misunderstand? What bothered you enough that you kept thinking about it? What competing concerns did you have to work through? What do you understand now that you couldn't have explained before? An essay could use sophisticated language without ever answering any of those more substantive questions. And sometimes the details that would help answer them actually get edited out by AI. An awkward but revealing observation then gets replaced with a broad statement about growth, or a complicated reaction becomes this perfectly neat lesson about perseverance or something along those lines. The result might be easier to read and more palatable quickly, but there's less to learn about you and especially about how you think. Now, none of these features prove that someone used AI. People write generic essays on their own too. Trust me, I've read many of them. The reason to fix them is that it just weakens the essay. I see this even when clients use an outline I gave them. I can often tell when they've just taken that outline, put it into ChatGPT and asked it, write me an essay about this. And that's also with an outline that we've already developed around their experiences after going through many rounds of brainstorming, after knowing their goals and knowing their experiences and what I want admissions to understand about them. So when they give it to ChatGPT, for instance, the draft now follows the structure, covers the points I want them to make, even the sentences sound really great, but the places where I wanted my client to think more deeply, to explain what bothered them, why they made that particular choice, or even how their understanding changed, that just gets filled in with language that sounds meaningful maybe, but it's actually telling me not that much. AI does a terrible job at coming up with insightful reflections when it's filling those gaps for the applicant. It can supply a plausible lesson, right? And it's going to be generic. That's how AI works. But that doesn't mean it's captured what this person actually wrestled with or came to understand. That's the work that Outline was asking them to do. So when AI fills those gaps, you can end up with something that looks like a really finished essay, a really well-written finished essay, while the most important questions are remaining unanswered and the most important insights are still missing. So when I say I can often recognize it, I'm not talking about catching just that particular word or an em dash or even a single sentence that follows the typical AI structure. I'm talking about knowing what we were trying to develop and seeing that the draft has written around it. And the applicant may not recognize the problem because the essay sounds so much better than their rough draft. It sounds finished before the thinking is finished. That's the key, the thinking. And a rough draft makes its problems visible. You can tell a transition isn't working. You know you haven't fully explained why an experience mattered. Or there's a paragraph where you're still trying to figure out, like, what point are you trying to make with this experience? But those gaps can be productive. They show you where you need to think more, and AI can cover those gaps in very polished, fluent language. So now the transition sounds smooth. They've figured out a way to connect that concluding paragraph. The experience has a nice lesson. Have you actually resolved the underlying problem? Part of this is also impatience. My clients go through many, many drafts to get to the level, to get to the standard that I expect my clients to be submitting. They often write upwards of 10 plus drafts. So if your personal statement isn't sounding or any of your essays aren't sounding where you want them to be after just a few drafts, don't immediately jump to AI thinking that's what you need. Let's take a hypothetical applicant who works as a paralegal and watches an attorney advise a client. An AI reflection might say something like, this experience revealed to me the importance of clear communication and it strengthened my commitment to advocating for others. Right. Really readable. Relevant to law. But what have I actually learned about this applicant? Very little. Maybe what actually interested this applicant was that the attorney couldn't give the client a definite answer. The law left room for uncertainty, and the client still needed to make a business decision, let's say. Maybe the client had assumed legal advice was primarily about identifying the correct But now watching that attorney-client conversation. Made them interested in how attorneys exercise judgment when several courses of action might be legally available. But then each of those, let's say, carries different consequences. And that's what really drew them in. Now we have something to explore, right? What did the applicant notice? Why did that question interest them? Did it change now how they approach their work or what they wanted to understand or what their future career path might look like? That material could reveal how the applicant thinks and why the work of a lawyer appeals to them uniquely. But if they accept the first polished sentence about communication advocacy that ChatGPT throws out, they're never going to develop that. Look how much more insightful the latter one is. Of course, this is an oversimplified example, right? But it's still the same point. You can give AI entirely accurate facts about your life and still end up with an essay whose interpretation doesn't meaningfully reflect your thinking. You've heard me say this so many times. Your thinking is the most important part of your essay. Having your experiences in the essay doesn't automatically make it a revealing or insightful personal statement. There's also a strategy problem. Your personal statement has a particular job within your application. And that job depends on your background, your experiences, your goals, and with the rest of your materials in your application already communicated. So maybe your resume shows a lot of substantial professional responsibility, but the reader still needs to understand why you want to leave that career you have now for law. Or maybe you have several experiences that seem unrelated and your essay needs to reveal the question or the interest connecting them. Maybe the story you're most eager to tell belongs better in a perspective statement while your personal statement actually should be developing another part of you, another part of your application. Those are all strategy decisions. Asking AI to just make this essay stronger doesn't establish that you've actually chosen the right topic from the beginning or that the essay is addressing what your application needs in terms of those strategic decisions. And it could just keep improving the sentences in a story that actually isn't even doing enough for you. So when I give the client an outline, I'm making decisions about what their application needs to communicate and identifying where we We might need more from them in particular to accomplish that. You know, having a paragraph where we needed an insight doesn't mean we now have that insight. That missing insight might have been the very thing we needed to distinguish this applicant, maybe to explain their motivation for law or connect experiences that they had that were seemingly unrelated and kind of all over the place. So that's why, for me, my concern goes well beyond whether someone can tell that you use ChatGPT And honestly, in a few years, I don't think this is going to be a problem for law schools. I'm concerned that you're going to submit something that feels finished, that sounds amazing, but it's not doing the work your application needs it to do for you. A school doesn't have to identify your essay as AI written to finish reading it without a clear reason to actually admit you. So what if you only used AI to edit? This is where a lot of applicants get worried. The ideas are mine, I promise. I wrote the draft. I just asked it to improve some of the wording. If that's the case, start with the school's rules. Don't assume all forms of editing or brainstorm assistance are actually permitted everywhere. Many schools expressly exclude all forms of AI use. And don't assume a tool's features are all equivalent just because they're available in the same program. But where assistance is allowed, pay attention to what's changing in your materials. If you write a paragraph and you ask AI to make it more sophisticated, then make it more compelling, then make it more concise, eventually it's going to be grammatically cleaner, it just will. But it's also going to be a lot less recognizable as something you personally would say. There's a lot less interviews in the law school admissions process than many other grad programs. Your essays need to reflect your voice. You need to come alive on the page. If you whitewash that through using AI, you lose a lot of what is going to get an admissions officer to personally connect with you across the page. Because ChatGPT will often remove the slightly unusual observation that actually made that paragraph interesting. It often is going to replace your uncertainty with some confident lesson you haven't possibly actually reached. It might even turn your specific interest in a problem into some broad claim about wanting to make an impact. Each change might seem small, but together they really, really can flatten the essay. So be careful. You don't need your personal statement to sound like a legal brief or a motivational speech. It doesn't need to be perfect. You need it to sound like a thoughtful, intelligent person explaining something that matters to them. That's why I often tell clients to start in their natural speaking voice. That's why we brainstorm a lot before. Explain the experience as if we were having a conversation. I tell them, tell me what happened, what you thought, why you're still trying to understand this. We can always work on clarity and polish later. Frankly, that's the easy stuff. First, we need something that is worth communicating, worth polishing. And if you ask AI whether your essay is good, don't mistake an encouraging response for a meaningful admissions assessment. A list of strengths and a few suggested edits doesn't establish that your essay, again, is serving your specific application well. But how do you know whether your essay still has substance? Here are a few questions I would ask. First, what does this essay reveal about me that a reader can't get from my resume? If the answer is mainly, you know, you're hardworking, compassionate, or obviously you're interested in law, you probably need to go deeper. Second question, where does the essay show my reasoning, show my thinking? Look for the places where you explain a choice. You examine an assumption you had, a work through attention, or reflect on an understanding you had. Describing an experience and announcing that it changed you doesn't actually show the reader how that change happened. Walk them through your internal thought process. Third, ask yourself, would I actually say this? Do I really believe this? If you read a sentence and think, hmm, that sounds impressive, but I'm not even really sure what it means, stop there, rewrite it in your own words. And finally, ask yourself, why is this material in my personal statement? What does it help admissions understand about me? Does it advance my essay's central point? Does it add to my overall application? Those questions will tell you much more about the strength of your essay than whether a detector gives it a reassuring score. And if you have already used AI heavily, go back to your original notes, your initial experiences, try explaining the central idea out loud maybe without looking at the generated draft. What do you naturally emphasize? Where do you hesitate? Where are the gaps What explanation do you give that never made it onto the AI page? That's often where you're going to find the material you still need to develop. So can law schools tell if AI wrote your personal statement? Sometimes they may suspect it. Detection definitely isn't certain, and you need to follow each school's rules regardless. But the question I most want you to ask is, what has this process done to the quality of my application? Has it helped me communicate my thinking more clearly, or has it just replaced the thinking with something that simply sounds good? A polished essay can still leave admissions with very little understanding of who you are, what interests you, or why they should choose you specifically. You definitely don't need an extraordinary life story. You don't need flowery, ornate language. You do need to develop the substance of your experience and make thoughtful decisions about how to present them. That work can be difficult. It can be uncomfortable. It's going to take longer than you want. I always tell my clients that. But it's also where personal statements become compelling. And if you want help getting started, I have a free personal statement guide with exercises to help you develop your material. I'll link it below along with information about working with me and my team. And if this helped you think differently about your essay, I would really really appreciate if you like this video and subscribe for more practical law school admissions advice.