AI at RIT This Spring

A Message from Christopher J. Collison, Director, AI Hub & Initiatives

Dear Colleagues,

In this Spring edition of the AI Hub Newsletter, you’ll find updates on RIT’s AI Foundry, student-led innovations like TutorBot, enterprise AI pilots, and upcoming events such as the Fram Lecture and the AI Summer Institute. We also highlight new internal tools in development and share how strategy, ethics, and infrastructure are evolving to support the thoughtful and practical use of AI across campus.

“The RIT AI Foundry is a university-supported innovation engine where students, faculty, and staff co-design practical AI tools to solve real challenges on campus. 
Built on a bottom-up, agile model, the Foundry enables secure, small-scale experimentation—developing minimally viable AI tools in local, risk-contained environments. When tools demonstrate real value, they are ready to scale through RIT’s evolving enterprise AI structure.”

Let’s start with a familiar dilemma: the classic chicken-and-egg conundrum. Do we purchase a robust enterprise AI system first and tell the world we are RIT AI-ready? Or do we work hard to develop grassroots tools and scale up what works? At RIT, we’re starting with the egg—and that egg is the RIT AI Foundry.

Think of the AI Foundry as a launchpad for practical, problem-solving AI: advising bots, research communication amplifiers, tutoring systems, and more. These tools begin life as local, secure, small-scale projects, built by student co-ops and faculty or staff innovators. They’re designed to test real value, feature by feature, without risk to sensitive data or campus operations.

In working with students, we emphasize not only the technical build but also understanding the customer and the problem they face. We guide students toward building a minimally viable product (MVP) in sandboxed, non-production environments, and help them think ahead: if this works, how would we scale it responsibly? That means considering risk, data security, and sustainability from day one.

Some tools will stay as eggs—useful experiments. Some will hatch into chickens. And if we get it right, a few might even grow into golden-egg-laying geese that redefine how we teach, learn, and work at RIT.

See It Live at Imagine RIT – April 26, 2025 (Gosnell 1365)

Experience the future of AI at RIT firsthand during:

To move from isolated grassroots tools to scalable, campus-wide systems, we need both robust infrastructure and thoughtful oversight. This spring, RIT’s ITS team launched Phase 1 of an enterprise AI pilot, focused on staff use of Microsoft Copilot, Google Gemini. We’re now preparing for Phase 2, which will expand to include faculty and students in the fall—enabling broader experimentation with generative AI in both academic and operational settings.

In parallel, we’re working closely with ITS and university leadership to ensure our systems are built on sound strategy, policy, and support. Our goals will be to:

• Define clear performance criteria for campus-wide AI tools and establish an Enterprise AI Selection Committee
• Develop and formalize ethical guidelines for responsible AI deployment
• Streamline risk assessment and compliance protocols for safe, scalable adoption
• Create a sustainable funding model to support student co-ops and long-term operational growth within the AI Foundry

We’ve begun forming AI Oversight and Enterprise Selection Committees that include representatives from ITS, Academic Affairs, compliance, faculty, staff, and students. These groups will help guide tool evaluation, recommend best practices, and ensure that our campus-wide AI use aligns with RIT core values such as ethics, accessibility, transparency, and accountability.

Already, the AI Foundry has developed bots ready for small-sample pilots at Commencement and Imagine RIT, all built on non-confidential data and designed for secure deployment through ChatGPT. These projects demonstrate what’s possible when we build AI solutions locally, ethically, and fast.

As our capabilities grow, we’re committed to scaling not just the technology, but also the governance that makes innovation responsible, inclusive and sustainable.

This semester, after approval through the Human Subjects Research Office, we completed a pilot study of TutorBot, RIT’s custom 24/7 AI tutor, with nearly 300 students in General Chemistry. Built through the AI Foundry and powered by RIT-designed instructional logic, TutorBot helps students tackle complex problems, even when faculty aren’t available.

Here’s what we learned:

• Students prefer live faculty in class (we agree!)
• TutorBot excels as a late-night homework companion, not a classroom replacement
• It supports critical thinking and step-by-step reasoning in problem solving

Try TutorBot at Imagine RIT in Gosnell 1365: TutorBot: RIT’s AI Tutor – Test It Yourself!

Jump into one of our interactive, AI-powered mini-lessons—from “Detective Thinking 101”, and “Myth & Meaning 101” to “AI Basics” and “Be Tomorrow’s Tiger.” See how TutorBot guides learners toward understanding through critical thinking and feedback-driven prompts.

Built at RIT using open-source materials, TutorBot is paving the way for affordable, accessible, and equitable tutoring—without publisher paywalls or subscriptions.

As we explore how to responsibly and effectively scale AI across campus, the AI Hub proposes the following Five-Tier Framework as a starting point for discussion. This classification is intended to help us think strategically about the different types of AI technologies emerging at RIT and how they might fit into our broader campus ecosystem:

• Tiers 1–2: Vendor-developed tools with limited customization, often domain-specific (e.g., advising software, accessibility tools). In these cases, we rely on external vendors to evolve their offerings with AI capability.
• Tier 3: Enterprise AI platforms like ChatGPT EDU, Microsoft Copilot, or Google Gemini are broad tools that could be licensed by the institution and made available campus-wide. They offer flexibility and integration but are not deeply tailored to RIT.
• Tier 4: Modular agents built by RIT power-users, trained on local datasets and often seeded by AI Foundry projects. These solutions operate within enterprise tools but bring deeper customization and security, enabling department- or function-specific impact. 
• Tier 5: Homegrown tools from the AI Foundry - our “eggs.” These are original solutions created at RIT to solve very specific institutional problems, ranging from tutoring systems to research support bots.

This framework is not a roadmap, but a lens to guide campus conversations. It can help teams reflect on what they might build, test, adopt, or scale, depending on their goals, resources, support structures, and readiness.

We welcome your feedback and invite you to help refine this model as we continue the university-wide conversation on integrating AI in thoughtful, meaningful ways.

From May 14–16, RIT will host a three-day Summer Institute on Teaching and Learning with a dedicated session on Friday morning 5/16 on the Future of AI in Higher Ed. This is co-organized with the Center for Teaching and Learning. Whether you’re just starting your AI journey or ready to build your own GPT, the Summer Institute offers something for everyone—from foundational sessions on what AI is and how it impacts education, to hands-on workshops on prompt engineering, custom GPT creation, and the responsible use of AI in classrooms. A keynote from Andrew White will spark big-picture thinking about the role of AI, creativity, and human agency in the digital age.

Faculty can explore the full agenda.

RIT’s grassroots, student-powered approach to AI was recently featured in the Rochester Business Journal. The article showcases how we’re building ethical, practical tools—like TutorBot and AdvisorBot—that directly serve our community. It’s a proud moment of recognition for our AI Hub, the AI Foundry, and the spirit of innovation at RIT.

Read the article

Whether you’re in the classroom, the lab, the office, or a support role, AI can help you work smarter, teach better, and think bigger.

• Explore resources at rit.edu/ai, for which updates are planned for May/June '25
• Join our workshops and pilot programs this spring
• Reach out to aihub@rit.edu with questions or ideas

Final Thought
At RIT, we don’t wait for perfection.
We build, we learn, and we grow—together.
The AI Foundry is our egg.
Let’s see how many chickens—and golden geese—we can raise.