A Quick Intro to Generative AI
Generative AI is a type of artificial intelligence that can create new content—like text, images, code, and more—based on patterns it’s learned from existing data. Think of it as a supercharged assistant that can help brainstorm, summarize, translate, and develop content.
The vast and uncharted terrain of AI and Large Language Models (LLMs)
With the arrival of Large Language Models that can analyze large amounts of files and data, build apps from just a few prompted sentences, or even work as autonomous agents on someone’s behalf, DePaul community members have access to incredibly powerful AI services. Although it can be quick and easy to crunch data or create something new with these AI services, there are important university safeguards to keep in mind.
When considering the use of any AI services, start with three critical questions:
Am I providing any protected DePaul data to an LLM?
Ease of use doesn’t change the importance of data protection. Most university data should never be uploaded or used with an LLM or AI-based tool. Please review the
Responsible Data Storage Matrix for data categories and specific examples.
Assuming I have complied with DePaul’s responsible data policy, am I creating an AI-based app, service, or tool that is interacting with any other members of the DePaul community?
The moment a DePaul student, faculty, or staff member considers creating a new technology service that interacts with other people in the DePaul community, this service must undergo a thorough design, security, and data review like any other product that is used in our technological environment.
Am I using an autonomous agent?
Autonomous agents are strictly prohibited at DePaul.
Some guiding principles for the use of AI tools at DePaul
Some of these tools can be used for personal use with great caution. Others shouldn’t ever be used by DePaul community members.
-
Feeding data into LLMs: Don’t provide any sensitive, highly sensitive, or internal restricted DePaul data (including in the aggregate) to any LLM platform. Specific examples of tools not allowed with DePaul data of any non-public classification include, but are not limited to, ChatGPT, Gemini, and Claude.
When in doubt, avoid supplying DePaul data to any LLM platform that isn’t Copilot Chat unless you know it is completely public information (i.e.
available on public-facing university bulletins or websites).
-
When to use Copilot Chat: Copilot Chat is the approved general purpose LLM platform at DePaul. DePaul has a contractual agreement with Microsoft so that we can be assured that data used in Copilot Chat remains private.
Even so, you should only use data categorized as public and internal non-restricted (i.e.
public websites, class information, employee job titles, etc.)
with Copilot Chat.
-
Unit specific AI-based tools: Your school, office, or department may already have access to powerful AI tools that have undergone security review. Check with your area about the data considerations and usage guidelines for these pre-approved tools. We expect more approved AI tools to be available soon.
-
App-building or “vibe-coding” tools: Assuming no protected data has been used, it is possible to build apps or tools for professional use if they are strictly used by the individual and not shared. Once an app or tool is shared with others, it requires a thorough review like any other technology application or service employed by the university. AI-based apps or tools created independently should never appear to be created or endorsed by the university, and the use of university tools to generate personal profit is prohibited.
-
Autonomous AI agents: A new wave of AI services are promising to automate your own job. In short, users permit an agent to work autonomously on their behalf. These autonomous agents like OpenClaw or Claude Cowork are explicitly prohibited even if it's only accessing your "own" account or data.
DePaul does not allow the use of autonomous agents of any kind.
Although we find ourselves in an era of increased productivity and possibility, the critical importance of data protection, security considerations, and responsible use of technology must be central to any new endeavor that utilizes artificial intelligence.
Always reach out before using AI tools
If you have any questions about the AI principles discussed in this newsletter, contact Information Services at
helpdesk.depaul.edu.
It’s best to use Copilot Chat for most business purposes at DePaul (if approved for use)
If you use Microsoft’s Copilot Chat by logging in with your BlueKey credentials, many of these common AI-related risks are addressed. Even when using Copilot Chat, make sure it is approved for your specific use.
Copilot Chat has numerous safeguards already built into the app. Microsoft ensures enterprise-grade security, privacy compliance, and no training on your data. It respects your policies and protects against most AI risks.
In light of these protections available through Copilot, we urge the DePaul community to utilize this AI tool rather than using other, potentially less secure entry points for AI, such as ChatGPT, Google Gemini, and the many other AI services that potentially put DePaul at risk, especially for data leakage.
Questions about using generative AI?
Before using any AI tools, all DePaul students, faculty, and staff are encouraged to consult with your schools, departments, instructors, or managers. Be transparent about your interest in using artificial intelligence and make sure it is approved through the appropriate authority or university entities.
-
DePaul students are encouraged to review course materials for each class and follow the guidance of their instructors. Never use AI tools without direct guidance and approval from instructors and school administrators.
-
DePaul faculty members are encouraged to learn more about AI in the classrooms through
this Teaching Commons website. Faculty can always reach out to the Center for Teaching & Learning about AI tools in the classroom.
-
DePaul staff members are encouraged to visit
this page for information about DePaul AI tools.
Questions about using a new AI tool?
Please
submit a ticket with the Help Desk before using or purchasing any new AI tools or software at DePaul. Our team can help you navigate through the potential risks and benefits of these new and powerful tools.
Looking Ahead
DePaul University’s approach reflects a broader shift in higher education: treating generative AI not as a threat, but as a tool to enrich learning. By combining ethical principles with practical strategies, faculty are empowered to lead the way in shaping how AI supports teaching and learning.
If you have questions about using or purchasing AI tools, please submit a ticket with the Help Desk.