While we don’t limit which tools you may use, how you can use AI tools—and what information you may enter into them—varies depending on the tool and the type of information involved.
NU ITS has three risk classifications of information, or what we call data: low, medium, and high.
Information that can generally be made available to the public without harm to the University, entities with an affiliatin to the University or to individuals. Low risk data is publicly available information (public directory information, university job postings, university policy, handbooks), results of personal opinion surveys, or your own personal, non-university data. Low Risk Data may be used in all AI tools.
Information that is not legally available to the public or can have moderate adverse impact on the organization. This includes student information (transcripts/grades, class schedules, advising records, etc.), NU IDs, employee information (demographic, marital status, country of citizenship, etc.)
Medium Risk Data may only be used in enterprise AI tools as long as it is unidentifiable and approved by NU ITS or other approved entities.
Examples of medium risk data that is not identifiable and may be used in Enterprise AI tools without review and approval include internal communications and memos that are not high risk, non-public reports, budgets, plans, financial information, and IT planning or audit logs.
Information that is confidential, restricted, or sensitive, or is protected by law, regulation, or sponsor requirement. This includes (staff identifiable information (social security numbers, driver’s license, passport, etc.), Student information (health, data on illegal behaviors, financial), or employee information (W-2, grievance information, disciplinary records, etc.). High Risk Data may not be used in free, paid consumer (non-enterprise) or enterprise tools).
There are three categories of tools you refer to when discussing AI tools across the NU system: free, paid consumer (non-enterprise), and enterprise. To get the most use out of your AI tools, NU ITS recommends you prioritize using enterprise tools for work.
Free tools are free accounts of any AI tool (ChatGPT, Gemini, Claude, etc.) that an individual creates themselves. Most free tools by default, train on your interactions, meaning any information you share could potentially become visible to individuals who should not have the information. While many tools have the option of turning off “train the model for everyone”, these tools offer little protection.
Paid consumer (non-enterprise) tools are AI tools that an individual pays for themselves and is not provided through a sanctioned enterprise environment. These tools may seem like they have increased privacy, and the common default setting that it does not train future models, but that cannot be guaranteed.
Enterprise tools are paid versions of any AI tool that are administered, owned and operated by the Office of the President or for which the University of Nebraska System is responsible and has made available to the University community. These tools offer enterprise grade security, ensuring your interactions with the AI tool are safe in our environment and do not train future models.
| Data Type | Data Examples | Free Tools | Paid Consumer | Enterprise Tools |
| Public Data | Anything publicly available on the web | Yes | Yes | Yes |
| Institutional (Non-Sensitive) Data | Public policies, handbooks, etc. | Yes | Yes | Yes |
| Student-Generated Learning Materials | Student lecture notes, reading summaries, etc. | Yes | Yes | Yes |
| Educational Materials | Readings, publisher, or faculty PowerPoints, etc. | No | No | Yes |
| Institutional (Medium Sensistive Data) | Internal memos, budgets, etc. | No | No | Yes |
| Anonymized Research Data | Identifiable information removed, other information generalized, etc. | No | No | Yes |
| Institutional (Sensitive) Data | Social security numbers, credit/debit cards, student health data, employees W-2, etc. | No | No | No |
| Raw Identifiable Research Data | Identifiable human subject data | No | No | No |
The University of Nebraska uses cookies to give you the best online experience. By clicking “I ACCEPT” and/or continuing to use this website without adjusting your browser settings, you accept the use of cookies.
