Elite College Students Rely on AI—But It's More Than Just Outsourcing Work

The Rapid Adoption of Generative AI Among College Students

A recent survey conducted by me and my colleague, economist Zara Contractor, revealed that over 80% of students at Middlebury College use generative AI for their coursework. This marks one of the fastest rates of technology adoption in recent history, far surpassing the 40% adoption rate among U.S. adults. The surge in usage occurred within less than two years after the public launch of ChatGPT.

Although the survey focused on a single institution, the results are consistent with other studies, offering a broader view of how AI is being used in higher education. Between December 2024 and February 2025, we surveyed more than 20% of Middlebury’s student body—634 students—to understand how they are utilizing artificial intelligence. Our findings were published in a working paper that has not yet undergone peer review.

The data challenges the widespread panic surrounding AI in academic settings and suggests that institutional policies should focus on how AI is used rather than whether it should be banned.

AI as a Learning Tool, Not Just a Shortcut

Contrary to alarming headlines suggesting that "ChatGPT has unraveled the entire academic project" or "AI cheating is getting worse," our research found that students primarily use AI to enhance their learning, not to avoid work. When asked about 10 different academic uses of AI—from explaining concepts and summarizing readings to proofreading, creating programming code, and even writing essays—explaining concepts was the most common use.

Students frequently described AI as an "on-demand tutor," especially useful when office hours weren’t available or when they needed immediate help late at night.

We categorized AI uses into two types: "augmentation," which enhances learning, and "automation," which produces work with minimal effort. We found that 61% of students who use AI employ it for augmentation, while 42% use it for automation tasks like writing essays or generating code.

Even when students used AI to automate tasks, they demonstrated judgment. In open-ended responses, many mentioned using AI during high-pressure times, such as exam week, or for low-stakes tasks like formatting bibliographies and drafting routine emails—not as a default approach to completing meaningful coursework.

A Global Trend?

Middlebury is a small liberal arts college with a relatively large portion of wealthy students. But what about other institutions? To explore this, we analyzed data from other researchers covering over 130 universities across more than 50 countries. The results mirrored our findings: globally, students who use AI tend to do so to augment their coursework rather than automate it.

But should we trust what students tell us about how they use AI? One concern with survey data is that students might underreport inappropriate uses, like essay writing, while overreporting legitimate ones, like getting explanations. To verify our findings, we compared our data with information from AI company Anthropic, which analyzed actual usage patterns from university email addresses linked to their chatbot, Claude AI.

Anthropic's data showed that "technical explanations" represent a major use, matching our finding that students most often use AI to explain concepts. Similarly, the company found that designing practice questions, editing essays, and summarizing materials account for a substantial share of student usage, aligning with our results.

In other words, our self-reported survey data matches actual AI conversation logs.

Why This Matters

As writer and academic Hua Hsu recently noted, "There are no reliable figures for how many American students use AI, just stories about how everyone is doing it." These stories often emphasize extreme examples, like a Columbia student who used AI "to cheat on nearly every assignment."

However, these anecdotes can conflate widespread adoption with universal cheating. Our data confirms that AI use is indeed widespread, but students primarily use it to enhance learning, not replace it. This distinction matters: By painting all AI use as cheating, alarmist coverage may normalize academic dishonesty, making responsible students feel naive for following rules when they believe "everyone else is doing it."

Moreover, this distorted picture provides biased information to university administrators, who need accurate data about actual student AI usage patterns to craft effective, evidence-based policies.

What’s Next?

Our findings suggest that extreme policies, such as blanket bans or unrestricted use, carry risks. Prohibitions may disproportionately harm students who benefit most from AI’s tutoring functions while creating unfair advantages for rule-breakers. On the other hand, unrestricted use could enable harmful automation practices that may undermine learning.

Instead of one-size-fits-all policies, our findings lead me to believe that institutions should focus on helping students distinguish beneficial AI uses from potentially harmful ones. Unfortunately, research on AI's actual learning impacts remains in its infancy—no studies I'm aware of have systematically tested how different types of AI use affect student learning outcomes, or whether AI impacts might be positive for some students but negative for others.

Until that evidence is available, everyone interested in how this technology is changing education must use their best judgment to determine how AI can foster learning.