For independent educational consultants (IECs), artificial intelligence (AI) and platforms like ChatGPT dangle an irresistible promise: dramatically reducing the hours spent reviewing and brainstorming student essays. Whether you are looking to grow your firm’s capacity or simply catch your breath during the October rush, AI seems to hold the key to transforming the countless hours spent on essay review. Yet integrating AI into the essay process raises both ethical and practical concerns.
The issue is not as simple as labeling AI good or bad. Many educators have formed strong opinions about AI without fully understanding what’s happening under the hood. This article goes beyond the surface-level debate: we examine how AI is currently being used in college applications, explore its underlying mechanics, reveal critical weaknesses that often go undiscussed, and ultimately provide a practical framework for leveraging AI ethically in your practice.
Examining the Landscape of AI Usage in College Applications
As with many new technologies, students are leading the charge. In the 2023–24 application cycle, one in three applicants used AI in some way for their college application essays.
Across the admissions aisle, universities are still catching up. Only 20 percent of the top 100 universities have established clear AI policies for college applications. However, those that have done so share a consistent message: use AI like a trusted adult. Just as you can guide students through brainstorming and offer feedback without writing their essays, AI can serve a similar role. A consensus is emerging: AI assistance is not considered cheating—when used appropriately.
But this raises a crucial question: Can an artificial system truly provide meaningful guidance on something as personal as a college essay? To answer this, we need to look beyond the hype and understand what AI actually is—and is not.
What’s Going On Under the Hood of AI
While ChatGPT is not the only AI platform, its widespread recognition makes it an ideal case study for understanding how AI systems work. Used by nearly a quarter of Americans each month, it is, at its core, a sophisticated chatbot—you ask a question, it responds with an answer. The important part is how it arrives at an answer.
AI is simply a program that uses a “majority rules” approach to answer questions.
ChatGPT was built in three steps:
- It began as a blank program.
- It was exposed to everything on the internet, including:
• Social media forums
• Newspaper articles - ChatGPT assumes everything it saw online was true and answers questions via “majority rules.”
As such, a great mental model to have when thinking about “what AI is doing under the hood” is: every time you ask ChatGPT a question, imagine it browsing Google, collecting 1,000 answers to your question, figuring out the most common answer, and spitting it back out to you.
While this “majority rules” process can work quite well for answering straightforward questions like “Why is the sky blue?” or “Tell me about IECA,” using it for college essay brainstorming or feedback raises one key concern.
AI’s Hidden Weakness That No One Talks About
While some online sources are genuinely valuable—like former admissions officers’ blogs—an overwhelming majority of sources are not. College Confidential threads, Reddit posts, and Quora forums number in the hundreds of thousands with students who, having gained admission to prestigious universities, believe they know the secret to success. However, as any IEC knows, no student truly understands why they were admitted—hence the well-meaning but misguided advice.
This is the crucial point about AI: it is not a magical oracle of admissions wisdom. Instead, it is essentially aggregating and averaging all this online advice, giving more weight to the most common perspectives. In other words, asking ChatGPT for college essay guidance is like polling thousands of College Confidential users at once—a prospect that should give any education professional pause.
Your Path Forward as an IEC
Students will likely be led astray by ChatGPT because they don’t know what makes a great college application essay. But you have the experience to parse good information from bad. And so, as an IEC, you are in the perfect position to leverage AI behind the scenes to help your student receive faster, high-quality feedback on their essays.
Broadly speaking, there are three ways to “get more” out of AI models like ChatGPT. The first two involve retraining ChatGPT itself—these methods are extremely effective because they involve fundamentally reteaching ChatGPT to focus only on data you specify. At Athena, we used both of these techniques to force ChatGPT to only learn from our collection of thousands of successful essays and proprietary knowledge from former admissions officers and IECs. However, these first two techniques require an AI background to implement and thus, may not be readily available to you.
The third technique is easily accessible to anyone reading this article—it is called “prompt engineering.” While less effective than retraining, it can still be a great way to get more out of ChatGPT than you do currently. Prompt engineering involves figuring out the specific words, phrases, and tone of writing to use when interacting with ChatGPT. Here are my five best prompt engineering tips:
- Give ChatGPT context. Explain that you are helping a student with their essay—the model tends to respond well when it understands that it is performing a “morally good” task.
- Enter in context about your student. The more ChatGPT knows about your student, the more personalized its responses will be.
- Start “new conversations” every now and then. Long conversations with ChatGPT tend to degrade its performance.
- Keep each instruction you give to ChatGPT as concise as possible. The more verbose you are, the worse its performance.
- Run the same essay through multiple times. You may receive different insights each time that can help you.
As a technology, AI has the potential to change how IECs like yourself fundamentally help students. And I’m truly excited for what the future holds.
Editor’s note: see IECA’s Considerations for the Ethical an Appropriate Use of AI for further guidance on using AI in your practice.
By Ajay Natarajan, President & CEO of Athena AI