Fifty-four doctoral students sat down to answer a deceptively simple question: what do you really think about ChatGPT? Their answers, collected by researchers at the University of Phoenix College of Doctoral Studies, reveal something more nuanced than the usual headlines about AI in classrooms suggest—that how students feel about artificial intelligence directly shapes how they use it, and that those feelings vary dramatically depending on what they study.
The research, led by Suchitra Veera of the University of Phoenix College of Business and Information Technology, examined the relationship between doctoral students' attitudes toward AI chatbots and their actual usage patterns. Published in the International Journal of AI in Pedagogy, Innovation, and Learning Futures, the quantitative study surveyed students enrolled at a private online university, asking them to consider questions around academic integrity, ethics, and the educational value of these tools. It's a question institutions increasingly need to answer as AI becomes woven into the fabric of higher learning.
The findings paint a clear picture: students who viewed AI chatbots favorably reported using ChatGPT more frequently. Those who believed chatbot-generated responses were superior to their own work also reported higher usage levels. And students who disagreed with the idea of prohibiting chatbot use altogether naturally gravitated toward using the tools more often. This might seem obvious—people use tools they trust—but the insight becomes powerful when paired with an unexpected finding: significant differences emerged across academic disciplines. A doctoral student in computer science approached ChatGPT differently than one in history or business. This variance matters enormously for institutions trying to craft policies that actually work.
What didn't matter? Gender. Researchers found no statistically significant differences in attitudes toward AI chatbots between male and female students, suggesting that concerns about gendered patterns of technology adoption may be less salient in graduate education than in other contexts.
"AI is rapidly reshaping how students approach research, writing and learning," Veera said. "Our research findings suggest that institutions should develop clear, discipline-sensitive guidance that supports ethical AI use while preserving academic integrity." That observation cuts to the heart of what makes this research valuable: not condemning AI tools or celebrating them uncritically, but acknowledging that students are already using them—thoughtfully, in many cases—and that institutions need to meet them there with intelligent policy, not blanket bans or blanket endorsement.
The team presented these findings at the 2025 Knowledge Without Boundaries Conference hosted by University of Phoenix, raising an important question for institutions scrambling to update their academic integrity policies: What does ethical AI use actually look like across different fields? A discipline-sensitive approach recognizes that AI's role in literary analysis differs fundamentally from its role in data science or business research. Rather than impose one-size-fits-all rules, institutions can now use this research as a foundation for developing guidance that speaks to students' actual needs and concerns within their specific programs.
This work suggests a path forward that treats students not as problems to be managed around AI, but as thoughtful users whose attitudes and experiences deserve serious institutional attention.