Supporting Generative AI for Faculty and Students

There can be little doubt that Generative Artificial Intelligence (GenAI) is making a profound impact on teaching, learning, and scholarship in higher education. A recent survey by the American Association of Colleges and Universities and Elon University’s Imagining the Digital Future Center observed that

• Student use of AI tools is nearly ubiquitous, but faculty use trails significantly behind.
• More than a third of higher education leaders surveyed perceive their institutions to be below average or far behind others in using GenAI tools.
• 91% expect GenAI will enhance and customize learning and 75% say it will improve students’ research skills

In January 2024, Keith W. McIntosh, Ed.D., VP for Information Services and CIO at the University of Richmond convened an AI Working Group to consider GenAI at UR. Members of the Provost’s Office, the Faculty Hub, Information Services, Boatwright Library, M. Saif Mehkari, Lauren Tilton, and Linda Boland began meeting regularly. Later in the year, Stephanie Spera and Marco Ortiz, director of the Weinstein Learning Center joined the group. The group was established in response to the rapid achievements and increasing integration of generative AI technologies within higher education. Recognizing the need for a structured approach at the University of Richmond to support faculty and students in navigating these changes, the group aimed to gather information and develop strategies that would enhance our academic community’s understanding and application of generative AI tools.

The group discussed the possibilities and challenges of generative AI at the University, with several  outcomes.

• A Generative AI website was created, with guidelines on the use of GenAI for students, faculty, and staff.
• Information Services updated the University Data Security Policy and the External Data Transfer Policy to include information on GenAI. A link to a list of AI courses in LinkedIn Learning was made available.
• Boatwright Library released a Research Guide on Generative Artificial Intelligence. The Library also released a guide to Citing AI.
• And as a result of the Faculty Hub’s Faculty Learning Community on Generative AI, a pilot was developed, called SpiderAI (note: SpiderAI is only available on the campus network or via the VPN). SpiderAI aggregates GenAI tools from OpenAI, Anthropic, and Google, providing access to students and faculty at no cost to them.

We sat down with Bobby Paxton, director of Web Services, to learn more about this initiative. Mehkari, Andrew Bell, and Paxton built and maintain SpiderAI.

What are the goals for SpiderAI?
SpiderAI provides equitable access to cutting-edge generative AI Large Language Models, or LLMs, facilitating AI literacy, encouraging exploration, and building a community to evaluate the benefits and risks of using AI in teaching and learning.

How does SpiderAI prevent the AIs from tracking people who submit prompts?
All requests are sent from a single source, so vendors only see requests are from the University of Richmond – they do not have access to individual user data. Additionally, our agreements with the vendors state that requests are retained for only 30 days and not used to train new models. However, user input does leave the University’s infrastructure, so we encourage users to review the Data Security Policy and ensure that they do not send information classified as classified, restricted, or official-use.

How does SpiderAI support equity for faculty and students?
SpiderAI offers centralized access to multiple LLMs—including those from OpenAI, Anthropic, and Google—through a University of Richmond–sponsored platform. It provides features like chat, image generation, text-to-speech, speech-to-text, and custom assistants. This eliminates the need for costly individual subscriptions. The Faculty Hub, WLC, and Library are also available for training and workshops, which we hope lowers the barrier for users in our community to explore and get comfortable using this new technology responsibly.

What has the usage of SpiderAI been like for the Fall 2024 Semester?
During Fall 2024, approximately 800 users made over 27,000 requests across various departments such as computer science, mathematics, English literature, foreign languages, social sciences, and economics; spanning teaching, research, and administrative functions.

Are there improvements in the works for Spring 2025?
Yes, we plan to expand the available models—including an on-premises Llama model and new offerings from OpenAI, Google, and more—and are exploring features such as no-code Retrieval Augmented Generation (RAG), API interfaces, video generation, speech-to-speech and user interface enhancements. Our roadmap for the year is still in development and subject to change based on resources and technological advancements.