EagleEye

Secure AI Platform NebulaOne is Now Available

After nearly a year of pilot programs and testing, today all UMW students, faculty, and staff have access to NebulaOne, a secure generative AI platform, at ai.umw.edu/nebulaone. Sign in with your UMW email address and password to access the platform.

NebulaOne gives the campus community access to premium models from several leading AI companies while keeping work inside the University’s own secure systems rather than sending it to a consumer service. Anything you enter into NebulaOne is contained within UMW’s Microsoft tenant and is not used to train or develop AI models.

The University pursued this platform for three reasons:

• Privacy – Work produced by the UMW community, as well as sensitive university data, should not be harvested by AI companies to use in the advancement of their own corporate goals. This platform offers a secure solution for those who want or need to use AI in their work to do so in a controlled, private environment.
• Equity – Access to capable AI models should not depend on who can afford a monthly subscription. This platform is available to everyone at UMW at no individual cost, allowing equal access for all to experiment with frontier technology, learn where and how AI tools fit into their work, and keep up with the rapid pace of AI development. 
• AI Literacy – A shared, supported environment lets us teach and study these tools together rather than leaving each person to figure them out alone.

A few things to keep in mind when using the platform:

• Highly sensitive data, as defined by University data classification policy, should never be entered into NebulaOne, nor any other AI platform. Protected/Internal Data (e.g. non-public memos, draft policies, internal reports, FERPA-protected records) may be entered into the platform with care and compliance with the AI Acceptable Use policy.
• Student use of AI in coursework continues to be at the discretion of each instructor. NebulaOne’s status as a “university authorized platform” simply means that it can be used for some work that would be inappropriate to share with consumer AI platforms. It does not mean that it is automatically approved for all courses or use-cases, nor that it is the sole option for all AI work at UMW. Consumer AI platforms (e.g. ChatGPT, Claude) may still be preferable for some use-cases, and Microsoft Copilot remains available as a simpler, but still secure, option for AI work. 
• Interactions with NebulaOne are not actively monitored by UMW staff, but the platform includes an automated moderation feature that flags dangerous or illegal activity for human review.
• All AI output should be considered a draft, not a final product. Verify what the platform produces before you share or act on it. AI models can be confidently wrong.

After logging in, users can interact with basic chatbots using NebulaOne’s OneChat feature, or dive into building more complex AI assistants using Personal and Shared Agents. To get started, see the NebulaOne Basics Guide, which walks through several features in more detail. 

To manage costs, a limited selection of models is available to all users, and usage of the platform is capped each month. If your work requires access to newer or more capable models, or you need additional monthly usage, contact the IT Help Desk.

NebulaOne is governed by UMW’s AI Working Group and a NebulaOne Governance Committee, working under the University’s AI Acceptable Use Policy. Questions, suggestions, issues, and bug reports are welcome and genuinely useful as the platform matures. All feedback can be submitted via the IT Help Desk.

About the platform: NebulaOne is built by Cloudforce, a Maryland technology firm specializing in secure AI deployment on Microsoft Azure. Microsoft named Cloudforce its 2025 Education Partner of the Year, and NebulaOne is in use at institutions including the University of Oxford, UCLA, the University of Maryland, and George Mason University.