AI Adoption and Sentiments Among Students: Survey Results

Artificial Intelligence (AI) in higher education remains an unsolved question. For students, balancing academic expectations, job market fears, and ethical concerns presents a particularly unique challenge. As such, many students feel drawn to use AI tools in some capacity in their work. At the start of Fall 2025 and 2026, HIS/POL students were given in-class surveys to understand how students are thinking about and utilizing tools in their classwork. Additionally, a focus group was hosted in Spring 2026 to gather information about specific AI sentiments.

The official Meredith College policy is “The use of artificial intelligence (AI) text generation tools may be considered a violation of the Honor Code. According to Honor Council Chair Laura Oppelaar, ‘27, “19 cases last year were specifically AI-related. This is a jump from 2024-2025, which was 13 cases.” AI tools should not be used in the completion of academic work unless an instructor specifically authorizes their use in a specific course or assignment. Students should refer to course syllabi and follow instructions by course instructors to determine appropriate use, if permitted,” according to the Undergraduate Student Handbook. In short, it’s up to individual professors to decide what is and isn’t appropriate for their class. Policies that vary even within departments can leave students confused when specific guidelines are not included in syllabi. Oppelaar said that AI usage in class work is “still detectable, and professors do detect it, but the lines are blurred between "outside" source and "plagiarism.” In the Spring 2026 focus group, students reported that while some professors specifically banned the use of AI, other professors simply instructed students not to use it, and some general education classes were reported as having no known AI policy.

Between 2025 and 2026, the percentage of students who self-reported using AI more than once a week rose from 49% to 53%. Students who self-reported not using AI at all fell from 27% to 16%, and over the course of the year, the percentage of students that felt confident or very confident using AI tools effectively rose from 79% to 85%. Between years, however, the percentage of students who were concerned about AI and felt that Meredith College should explicitly disallow it held steady at about 28%. Students who did use AI in Fall 2026 predominantly reported using it for generating study materials from notes, getting feedback, summarizing readings, and researching background information. In the focus group, students stated that they chose to use AI rather than ask professors either due to lack of access like not being able to come to office hours or needing an immediate response, or out of embarrassment for asking “dumb questions.” Only 3% of students self-reported using it to generate text.

Most students in Fall 2026, however, reported some awareness of concerns about AI and copyright infringement, environmental and energy impact, job loss, human rights and privacy violations, and misinformation. Not a single student surveyed said they were unaware or just slightly aware of the problems surrounding AI and misinformation. The percentage of students that were confident or very confident that they could identify AI generated content rose from 58% to 71%. Conversely, the percentage of students who said that they were confident that AI generated information was accurate fell from 17% to 8%, and no students felt very confident that AI generated accurate information. As a result, 86% of students self-reported that they fact checked any AI generated outputs before using them in coursework. 

Regarding the college’s policy, students are split. In both surveys, about half of students felt that Meredith College should continue to revise and adapt its AI policies in order to decide where and how AI should be integrated. In the Fall 2026 survey, 62% of respondents agreed or strongly agreed that AI should not be adapted into coursework, and 64% agreed or strongly agreed that it should. 

So the AI question remains: how should higher education respond? Although AI has been linked with productivity gains and is well adopted by students, the American Psychological Association (APA) found that students who used AI for tutoring saw grade improvements, but “performed no better than the control group once the tool was removed.” Students in the control group reflected this sentiment; some were concerned about “bypassing” the process of making study materials themselves or worried that they could become overly reliant on AI tools. A study from China of students in grades 7 to 12 published earlier this year by the Centre for Economic Policy Research (CEPR) found that while grades rose by 18%, monthly exam scores fell by 20%. The effect of AI on neuroplasticity and deep learning is still in its infancy, and although students are adopting it more and more, it calls into question what the long term neurological implications of AI usage in education truly are. For now, the question is up to professors to either embrace the productivity bump or stick to traditional learning.