Artificial Intelligence (AI) at Virginia Tech
Overview
Generative Artificial Intelligence models are quickly becoming part of everyday life in higher education. Learning to navigate AI integration and use by faculty, staff, and students will be an ongoing process. In this early stage, there are several general guidelines that the University community should keep in mind:
- Understand the security concerns involved in using AI tools.
- Explore and experiment with AI carefully.
- Do not discuss or input sensitive or protected data (PII, FERPA) when interacting with AI tools.
- Thoroughly review AI output against known sources to assure the information generated is accurate and correct, just as you would sources like Wikipedia.
- Do not give AI large and complex tasks.
- Do not install AI tools or plug-ins on university owned computers without discussing with your departmental IT staff.
TLOS has put together an excellent guide for Instructors: Considering Generative AI at Virginia Tech. There is even a GenAI community list you can join for updates.
Specific Use Cases
Administrative Use
Things to keep in mind for Administrative Use of AI:
- Outputs from generative AI are based off of existing data from a myriad of sources, which may be inaccurate.
- It is not always clear how specific outputs are created by generative AI (even by the creators of the model producing the output), or where the AI is sourcing the data that it has used to formulate its outputs. This could make it very difficult to determine accuracy, relevancy, and possible intellectual property violations of a specific output, without potentially creating many hours of additional research.
- Data that you submit to a generative AI model may become part of the model's database; using an AI tool may contribute to its development and the continued use of your inputs in the form of data that it pulls from to formulate new outputs for anyone else that is using the model. Do not enter any confidential, protected, or sensitive data into an AI system.
- MS Copilot: Understanding Microsoft Copilot with Data Protection at Virginia Tech
Uses:
- Scheduling and Calendars
- Drafting email correspondence
- Customer support Chatbots
- Filtering/organizing large datasets
- Code creation for website function/design
- Meeting transcription/recording/summarization
- Video/Audio editing
Research Use
- Research Security: Guidance for using artificial intelligence during research activities
Teaching and Learning Use
- Generative AI tools have proliferated to a point where students and faculty will be using them extensively. Like any tool they can be used for legitimate and illegitimate purposes.
- As instructors, it's up to you to determine how you will handle student use of AI and hold students accountable to the honor code.
- It is important to note that AI 'checkers' can generate false positives. Students have access to the same AI detection software as instructors, so it's very easy to disguise the use of AI.
- The following article goes into detail about the educational ramifications of AI use: Considering Generative AI and ChatGPT at Virginia Tech
AI Resources
- Considering Generative AI and ChatGPT at Virginia Tech
- Virginia Tech DCSS AI Overview/Update
- Research Security: Guidance for using artificial intelligence during research activities
- MS Copilot: Understanding Microsoft Copilot with Data Protection at Virginia Tech
- Library AI Guide: Virginia Tech Library AI Guidance & Policies
- TLOS - Intro to AI course.