Frequently Asked Questions

TerrierGPT | Sustainability | BU’s Investment in AI | AI and Me

Terrier GPT & Other AI Tools

What exactly is TerrierGPT?  Did the University build its own AI model?

No, TerrierGPT is not a new AI model; it’s a gateway for BU faculty, staff, and students to have equitable access to leading models — such as ChatGPT from OpenAI, Claude from Anthropic, Google Gemini, Meta Llama, and more — in a secure environment.  With TerrierGPT, you automatically get access to these foundation models without having to create separate accounts or share personal information. 

Isn’t it true that TerrierGPT hallucinates, gets basic facts wrong and doesn’t know current events?

You’re right.  Using GenAI tools today comes with its own unique set of challenges, and the outputs from AI models must be reviewed and fact-checked by a human-in-the-loop with a critical eye.  The training data of any AI model also reflects a snapshot of the world at a point in time in the past.  If your query requires knowledge of the past few months, a web search is probably a better starting point for now.  We expect integrated web search to be available soon.  You can also augment models with additional knowledge by uploading documents, for example.

I use ChatGPT. Why would I need or want TerrierGPT?

When you use ChatGPT or any other foundation model, your data is not protected.  Providers train their models on your queries and interactions.  Any private information or intellectual property you divulge may be revealed in response to a user query by a future version of their model.  With TerrierGPT, your data stays within the enterprise and is not used for training.  Your data is also kept private – BU IS&T provides the same privacy guarantees with TerrierGPT queries as they do with your BU email. BU is not using any of the conversations in TerrierGPT, neither to train its own models nor to furnish those conversations back to model providers.

I need to use the most cutting-edge models for my research.  Does TerrierGPT have me covered?

Probably not.  While we are constantly updating the supported models, research that either requires premium-level access to the very best models or involves programmatic usage through API calls are not use cases currently supported by TerrierGPT.  Please contact us at the IT Help Desk (ithelp@bu.edu) if you’d like to discuss your options. 

Where do I get answers for technical questions about TerrierGPT?

BU IS&T maintains its own FAQ about TerrierGPT that you can consult.   But the best place to get answers is within TerrierGPT itself.  The TerrierGPT assistant—available from the models drop-down under “Agents” once you log in—is an AI chatbot designed to answer your questions about TerrierGPT and AI resources at BU more broadly.

Are there other AI tools AIDA recommends in addition to TerrierGPT?

Yes!  Please visit our AI Tools page for a list of AI tools we recommend for use cases that go beyond LLMs. These tools have been reviewed by our experts and are clearly labeled by how much data protection they offer.

Sustainability

Doesn’t generative AI use huge amounts of energy and harm the environment?
Like many high-performance computing tools, generative AI models require significant computing power, which means higher electricity usage and associated carbon emissions.  Model training is especially compute-intensive.  However:

  • While training a GenAI model consumes an immense amount of energy, TerrierGPT is not a new model — TerrierGPT leverages existing LLMs available on the market. This approach minimizes resource use while still enabling advanced AI capabilities.
  • Queries do consume energy and aggregate consumption at the scale of BU is indeed significant, but it is helpful to be mindful of the relative consumption costs.
Source:  “The Environmental Impact of AI in Education: Cutting Through the Noise”, Anthology white paper, 2025. (Anthology is an EdTech firm owned by Blackboard.)  https://www.anthology.com/paper/the-environmental-impact-of-ai-in-education.
  • We encourage mindful use. AI isn’t meant to be used as a default solution, but to assist you when it adds clear value.
  • Sustainability in AI is a growing area of research among our faculty. Several faculty are actively exploring methods to reduce the environmental footprint associated with developing and operating LLMs. Their work is helping to shape a future in which AI innovation aligns with principles of environmental responsibility.
  • AIDA is partnering with the BU Office of Sustainability to align our initiatives with the goals of Sustainability 2030.

BU’s Investment in AI

Why invest in AI tools now, when resources are so limited?

Our AI investment is modest compared to other expenses, and it’s intended to provide equitable access to advanced tools students will encounter in their careers and help faculty and staff work more efficiently, freeing more time for creative work and direct student engagement. By providing TerrierGPT to the entire BU community, we aim to avoid unnecessary costs associated with external AI vendors and premium subscriptions. Additionally, TerrierGPT was implemented to ensure data security, recognizing that popular generative AI tools previously used by faculty, staff, and students do not necessarily safeguard BU data.

More broadly: while the University continues to face financial headwinds, we also continue to make strategic investments that align with our mission, prioritize people, and support our students. One key mission is to create holistic citizens ready to engage in life-long learning and service. It is becoming increasingly clear that, regardless of discipline, the workforce of tomorrow will require basic GenAI literacy and the ability to use GenAI tools to boost efficiency and creative problem solving. BU has an obligation to embed these skills in our graduates and in a way that enhances their career prospects, productivity, and ability to think critically and engage deeply in a complex, changing world.

AI and Me

I’m a faculty member. Do you have advice for setting effective AI policies in my classroom?

Yes.  BU faculty have broad latitude in setting their AI policies that best align with and facilitate their pedagogical approach.  But setting policies involves tradeoffs and considerable nuance. Please visit our Generative AI Guidelines for Classroom Use page for specific recommendations drawn from the shared experiences of BU faculty.  Your unit or college may also have recommended AI policy settings.  This topic will also be covered in our forthcoming primer on AI for faculty.  More information will be released soon.

I’m a student. How do I know what’s allowed and what’s not when it comes to AI?

AI is rapidly evolving, and BU faculty members may freely decide how to set the AI policies for each of their courses, within broad limits established by the BU Academic Conduct Code.  Variation across courses is the norm, not the exception.  Students are encouraged to review the course policies and consult their instructors for guidance.  Please visit our Generative AI Guidelines for Students page for specific topics you should consider.

Could over-reliance on AI diminish my critical thinking skills and ability to learn?

Yes, this is a legitimate concern.  Over-reliance could reduce deep thinking and hinder development of problem-solving skills if one turns to AI by default and uses it as a crutch. That’s why we recommend critical engagement with AI: understanding when to use it, when not to, and carefully checking its work.  Our forthcoming AI@BU course for students addresses this important topic in depth and helps you develop a personal code of conduct while using AI.  More information will be released soon.

How can I further enhance my understanding of AI?

  • AIDA Website & Newsletter:
 All faculty and staff have full access to AIDA’s website and newsletter, which provide regularly updated news, guidance, tools, educational materials, and event announcements designed for the entire university community.
  • IS&T Training for Generative AI Tools: Our Information Services & Technology (IS&T) department offers dedicated training sessions for generative AI tools tailored for staff. You can see available offerings and register for upcoming sessions here: IS&T Training for Generative AI Tools.
  • Terrier eDevelopment: BU’s Terrier eDevelopment platform includes a variety of AI-related courses and resources specifically curated for staff. These flexible, self-paced modules cover practical AI applications and concepts relevant to administrative and operational roles.
  • AIDA Symposiums: AIDA regularly hosts symposiums focused on AI, open to staff and faculty alike. These gatherings provide opportunities for learning, discussion, and engagement with leading AI practitioners and BU’s internal experts. More information and current schedule can be found on AIDA’s News and Events page.

How is AI being used in the classroom at BU?

BU faculty are using AI in a range of creative ways that support our academic mission, including to support personalized learning (e.g., tutoring-style Q&A), to demonstrate and critique AI’s limitations, and to teach students how to think critically in a GenAI world. Each faculty member decides how (and whether) to integrate AI into their courses, and AI use is always governed by our academic integrity policies. Visit our Classroom Resources page for real-life examples of how faculty are using AI in their classrooms.