Frequently Asked Questions
For complete details on data classification levels, see the UMB Data Classification Policy.
Only if your research data is completely public and non-sensitive. Free versions are Level 0 tools and must never contain:
- Patient data or health information
- Student records
- Unpublished research findings
- Any confidential or proprietary information
For research with sensitive data, use Level 2 approved tools like M365 Copilot or NebulaOne.
ChatGPT Business, Claude Team, and Gemini Business are currently in the procurement and setup phase. They will be available soon with centrally managed accounts distributed through the Software Licensing Office. All accounts must use @umaryland.edu email addresses.
Check back regularly for updates or contact IT Security for more information.
Submit an IT-PCS Software Request Form. IT Security and Compliance will review the tool and determine if it meets UMB's security and data protection requirements.
Some AI tools are prohibited because they do not meet UMB's security and compliance requirements. Common reasons include:
- Storing data on external systems UMB cannot control
- Insufficient data protection guarantees or lack of institutional agreements
- Unpredictable behavior that violates privacy or consent requirements
- Non-compliance with institutional data protection policies
- Legal or regulatory concerns
Before using any AI tool: Check the approved list above. If a tool is not listed, submit an IT-PCS Software Request Form for evaluation before use.
Using AI tools with inappropriate data classifications can result in:
- Data breaches and exposure of sensitive information
- HIPAA, FERPA, or other regulatory violations
- Disciplinary action
- Legal consequences for the university
When in doubt, contact IT Security before using an AI tool with any university data.
Contact
Help Desk
Security & Compliance
For AI security questions, data classification guidance, or compliance concerns
When to Contact:
- Questions about data classification levels
- AI security concerns or incidents
- Compliance requirements for AI tools
- Requesting approval for new AI tools