Home  »  Getting Your Degree   »   MIT Report Addresses AI in Higher Education

MIT Report Addresses AI in Higher Education

Higher education cannot and should not use uniform rules or blanket policies when it comes to regulating artificial intelligence in classrooms. That’s according to a report by the Massachusetts Institute of Technology, published by its Committee on AI Use in Teaching, Learning, and Research Training.

The committee examined the effects of machine learning software and AI in general on university coursework. Their findings are relevant for both instructors and students.

MIT Report Addresses AI in Higher Education

The introduction to this report includes the following:

“…what we learned as a group quickly convinced us that the Institute community, particularly the faculty, must tackle a set of deeper questions about the structure, meaning, and value” of education “…in an era in which AI is one of several factors” that inform higher education in the 21st century.

The report notes that students “…use AI frequently and pervasively – with strongly mixed feelings, from curiosity, creative inspiration, and gratitude to resignation, concern, and anxiety. Instructors’ attitudes range from enthusiastic exploration and growing reliance on AI to skepticism, suspicion, and ‘AI refusal’– and there’s a widespread desire to share experiences, ideas and techniques.”

>> Find VA-approved schools — please try our School Finder tool.

AI at MIT and Beyond

The details of this report focus on the activity at MIT, but the findings can be interpreted in the larger higher education context.

According to MIT, “Through five intense months of meetings, research, and outreach across the MIT community, the Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training sought to understand the role of generative AI in the life and educational mission of the Institute and recommend how to navigate its challenges and opportunities.”

Generative tools can produce answers for homework or write working computer code that appears human-generated. AI’s ability to create and interact, according to the study, makes it necessary for academic departments to establish new AI-related policies that are specific to the course of study rather than blanket rules.

AI text generators produce responses to assignment prompts within seconds, making traditional homework models seem obsolete. They now fail to show whether a student understands the assigned material, at least hypothetically, because AI can step in to fill a knowledge gap.

AI and the Illusion of Learning

The report notes that “getting the right answer from a chatbot can create the illusion of learning” without teaching core concepts, and when students rely on AI to solve problems, they bypass the effort needed to learn the material.

To address this breakdown in homework reliability, the committee advised faculty to reform grading practices. With AI, it’s more important for instructors to evaluate the learning process rather than just finished coursework. The committee also urges faculty to replace take-home problem sets with interactive evaluations.

>> In order to find VA-approved schools, please try our School Finder tool.

Recommendations

Recommended test methods in the age of AI may include oral examinations, live demonstrations, and project portfolios compiled over the course of a semester. When students explain their reasoning aloud, teachers can gauge their learning in real time, which may be one reason why the report advises instructors to clearly define the role of artificial intelligence on all course descriptions.

The guidelines should include rules on when AI use is permitted and when students must complete work unaided.

The committee warns against using so-called AI detectors to catch students “cheating” with artificial intelligence. Why? Common AI detection tools produce false positives, misidentifying student work as machine-generated text and creating mistrust between instructors and learners.

Colleges must, according to the report, instead train students in ethical software use and design assignments around tasks that automated systems cannot perform on their own.

Using Machine Learning on Campus

40% of surveyed members of the campus community reported regularly using machine learning software. Students turned to these systems to debug code, brainstorm essay topics, and summarize research papers. But only 23 percent of respondents expressed optimism about long-term shifts in education.

Faculty members voiced concern about skill erosion, while students questioned whether their credentials would prepare them for professional fields altered by automation. Another problem? Equity issues, which form a central part of the committee’s findings.

Access to paid AI subscriptions creates resource divisions within student bodies. Commercial companies place advanced model parameters and faster processing speeds behind monthly paywalls, giving wealthier students access to superior study aids than those who can’t afford the fees.

At the graduate level, the report outlines challenges for research training and thesis preparation.

Doctoral candidates and postdoctoral researchers use language models for literature reviews and laboratory data, just to name a few uses. The committee’s report says AI invents citations and misrepresents scientific data, which is a real problem for researchers who don’t anticipate it. The authors stress that humans, not software, must review the accuracy of published scientific claims; they cannot blame AI tools for errors in procedure, formatting, or fact.

The report concluded that artificial intelligence will continue to alter campus education and workforce expectations, requiring universities to adjust degree requirements as the technology disrupts higher education.

>> In order to find VA-approved schools, please try our School Finder tool.