TLDR: A recent study by Duke’s Center for Applied Research and Design in Transformative Education (CARADITE) reveals that while Duke University students frequently use generative AI tools, many express skepticism regarding their accuracy and call for more guidance from instructors. Despite these reservations, a significant majority believe AI will lead to more personalized learning and will be crucial in their future careers.
Durham, NC – October 17, 2025 – As Duke University continues its embrace of artificial intelligence in the learning process, a new research study indicates that students are approaching the technology with a blend of frequent use, critical skepticism, and measured optimism. The study, titled ‘Gauging Duke Students’ Perspectives on Generative AI,’ was conducted by Duke’s Center for Applied Research and Design in Transformative Education (CARADITE) and collected data throughout April 2025 from 104 undergraduate and graduate students across more than 40 fields of study. The findings were recently published to coincide with the new academic year, during which all Duke undergraduate students gained no-cost access to ChatGPT.
The research highlights a pervasive integration of AI into students’ academic lives. Students routinely utilize AI as a writing assistant, a problem-solving tool, and a thought partner. However, this frequent use is tempered by significant concerns about the technology’s reliability. A striking 75% of students reported that AI often provides inaccurate answers to prompts, and 62% noted that AI tends to oversimplify content, potentially hindering deeper understanding. One student articulated this caution, stating, ‘If I use AI to guide me in problem solving for math or physics, I have to double check the work and correct the AI. I use it mostly as a guide or a starting point, not to give me finished answers to skip my work.’ Another student expressed a more direct concern: ‘It is inaccurate, and I am worried about the consequences.’
Alongside these reservations, students are actively seeking more educational guidance from their instructors on the ethical and effective use of AI. As one student emphasized, ‘As an educator, it is your responsibility to educate your students about the ethics of AI, as well as its unreliability.’ The study underscores the need for faculty, staff, and administrators to better understand and support students through this technological shift.
Also Read:
- Harvard Pioneers Adaptive AI Frameworks for Modern Pedagogy
- Study Uncovers Generative AI’s Role in Enhancing Human Creativity and Reducing Cognitive Burden
Despite the expressed concerns, the overall sentiment among Duke students leans towards optimism regarding AI’s long-term impact on education and their professional futures. A substantial 80% of students believe that AI will enable more personalized learning within the next five years. Furthermore, over half (56%) anticipate that AI will play a significant role in shaping the industries they plan to enter. This suggests that while students are cautious about current AI limitations, they recognize its transformative potential and inevitability in their educational and career trajectories.
