This is an early release of this KB article and content is subject to change. Please refer back to this article occasionally for updated guidance.
Anthropic Claude is an AI-powered assistant designed to help users write, analyze information, summarize documents, answer questions, generate content, and support research and administrative work.
Claude is available to UVM employees through a university-managed team license that includes privacy and security protections. Like other enterprise AI services approved by UVM, data submitted to Claude through the UVM-managed service is protected under UVM’s agreement with the vendor and is not used to train public AI models. A Claude license also grants access to Claude Code.
UVM’s preferred and supported AI platform is Microsoft Copilot. Copilot and M365 Copilot are licensed in the university’s Microsoft 365 ecosystem and provides institutionally approved access to AI capabilities. Faculty and staff should evaluate whether Copilot meets their needs before requesting access to Claude. Claude may be appropriate for specialized use cases where its capabilities provide a clear advantage, but the use of additional AI platforms should be based on a demonstrated need beyond what Copilot can provide.
Accessing Claude and Claude Code
Claude and Claude Code access is managed through a UVM Team License and is provided through a chargeback model where users and departments pay for the license(s) they consume. This model allows UVM to take advantage of security and privacy features under the Team terms. Current Team licenses are available at $20/month for a standard seat and $100/month for a premium seat.
To request a Claude license, fill out this form: https://forms.cloud.microsoft/r/udSv4sJRup
Why Use Claude?
Microsoft Copilot is included with every UVM Microsoft 365 account and is an excellent choice for many day-to-day tasks, especially when working within the Microsoft ecosystem.
Claude may be a better choice when:
- Working with long documents, reports, policies, research papers, or meeting notes.
- Synthesizing information from multiple sources.
- Creating detailed drafts, analyses, or structured reports.
- Performing literature review and research support activities.
- Iterating on complex writing projects.
- Exploring complex questions that benefit from detailed reasoning and explanation.
Privacy and Data Protection
Use Claude through your UVM-provided account.
Personal Claude accounts are not covered by UVM’s enterprise agreement and should not be used with UVM non-public information.
When using the UVM enterprise license:
- Low, Medium, and High Risk data may be used in accordance with UVM policy and the Data Risk Classification Matrix.
- Users remain responsible for protecting regulated data and complying with university policies and legal requirements.
- AI-generated results should always be reviewed for accuracy and appropriateness before use.
Example Uses
Administrative and Academic Work
- Drafting emails, reports, and communications.
- Summarizing lengthy documents.
- Developing meeting agendas and action plans.
- Analyzing survey results and qualitative feedback.
- Preparing presentations and training materials.
Research Applications
- Summarizing research articles.
- Assisting with literature reviews.
- Identifying themes across collections of documents.
- Assisting with research communications and outreach materials.
Verify AI Output
Generative AI systems can occasionally produce inaccurate or misleading information, sometimes called “hallucinations.”
Before relying on AI-generated content:
- Verify facts and citations.
- Review calculations and analyses.
- Confirm policy, legal, regulatory, or compliance information.
- Apply human judgment before publishing or sharing results.
Additional Resources
Agentic AI Dos and Do Nots KB Article
Agentic AI Guidance KB Article
Claude Code Permissions KB Article