Duke Develops AI Education Program for Academic Leaders

Taking notes, doing their homework, presenting their projects—it’s been back to school for many of Duke’s academic leaders this year as they have worked to deepen their understanding of artificial intelligence, or AI. 

Duke’s AI Education Program for Academic Leaders is a yearlong, project-based initiative for the Provost’s executive leadership team, which includes deans, vice provosts, and other senior leaders. Through a structured sequence of workshops and training sessions, academic leaders gained hands-on experience with Duke’s AI tools in addition to a high level of knowledge about the technology and its implications for the university. Leaders also developed their own AI projects, identifying a particular application under their administrative purview and testing whether AI could streamline or improve their workflow. 

The program was developed and led by Provost Alec D. Gallimore in close partnership with Tracy Futhey, vice president for information technology and chief information officer; Matt Hirschey, associate professor of medicine and director of the Center for Computational Thinking; and Joseph Salem, Rita DiGiallonardo Holloway University Librarian and vice provost for library affairs.

Image

“Our academic leaders are helping Duke navigate through a moment of rapid technological change,” Provost Gallimore said. “Our goal for this program was to give this team of leaders a customized, in-depth learning experience that would equip them with both the technical and conceptual understanding of AI to make strategic decisions about how to deploy it – or not deploy it – in their college, schools and units. As students, staff and faculty experiment with the potential of AI, it is vital that leaders and decision-makers have a sophisticated and nuanced grounding in its capabilities and limitations as well as the many ethical considerations.” 

The education program for academic leaders is connected to Duke’s broader engagement with AI, which aims to build literacy in the technology, use AI to innovate across the university, and support its responsible use. Duke provides secure, institutionally licensed access to tools and applications, ensuring that all students, faculty, and staff can safely experiment with AI. 

“At a time when many students and faculty are racing to use AI, we academic leaders need to keep up—and stay ahead,” said Jerry Lynch, Vinik Dean of the Pratt School of Engineering. “This educational program was an efficient way for Duke’s academic leaders to gain practical experience with AI, identify ways to use it, and anticipate future issues.” 

The training sessions covered best practices for prompting and applying AI, the distinction between open-source and proprietary models, emerging reasoning capabilities, and data privacy practices using Duke’s secure systems. By the end of the academic year, the program aims to establish a shared baseline of AI competence among academic leaders, a portfolio of tested AI use cases, and a sustainable culture of innovation through the responsible use of AI.

“Although the academic leadership team came in with varying degrees of experience with AI, they were uniformly curious and eager to learn,” said Hirschey, who led many of the sessions. “I enjoyed helping them get their minds around this rapidly changing technology and refine their thinking about how these tools can be used in teaching, research, and operations in the years to come.” 

Image

A special session in October featured a conversation with Sarah Friar, CFO of OpenAI, and Ronnie Chatterji, Mark Burgess & Lisa Benson-Burgess Distinguished Professor of Business and Public Policy and chief economist of OpenAI, that highlighted the company’s partnership with Duke and explored how universities can harness AI responsibly.

“AI competency is quickly becoming a core leadership skill,” said Lori Bennear, Stanback Dean of the Nicholas School of the Environment. “I got so much out of the experience that I copied the model for our school. Every unit now has an AI lead working through the training and developing a concrete project to improve efficiency. The goal is simple: work smarter, not harder, and unlock the innovation AI makes possible when it’s used well.”

After developing their projects, the academic leaders shared early results at a session in December. For example, one dean used AI to survey university policies and governance documents and identify outdated, redundant, or conflicting policies. Other leaders used AI to organize documents, update course materials, or summarize knowledge repositories in ways that reduced staff workload, improved efficiency and accuracy, and extended their capabilities. 

“It has been a valuable experience to work on the AI project with my team,” said Suzanne Barbour, dean of The Graduate School and vice provost for graduate education. “We look forward to leveraging AI to mine graduate student data. Much of that data is housed and managed on various platforms across the university, and AI can pull the data together rapidly and run analysis for dashboards and other reporting efforts.” 

Administrators will be invited to discuss their experiences with the training program—including reports on implementing their AI projects—at the Duke AI Summit in May.