Security and Data Privacy Statement

Google Workspace has long-standing, robust privacy commitments to protect user data, and the introduction of generative AI, like Gemini, reaffirms their importance. These commitments ensure that generative AI does not alter the foundational privacy protections that give you choice and control over your data.

• Your interactions with Gemini stay within your organization, and Gemini does not share your content outside your organization without your permission.
• Crucially, your content is not human-reviewed or used for Generative AI model training outside your domain without your permission.

Data Use and Access for Gemini and NotebookLM

• Gemini is designed to access relevant Google Workspace content based on your prompt and that you have permission to access. It uses your prompt, relevant Workspace content, and webpage context to generate a response.
• For NotebookLM, users can upload sources from Workspace only if they have permission to access them
• When generated responses are inserted into an email or document, Google automatically applies your existing data protection controls.

Data Classification

Google Gemini has data classification moderate, which is approved for up to 'Level 2' data.

To learn more about the classification, please review IT-2 University of Maryland Data Classification Standard.

Introduction

As part of the Google Workspace suite, the University of Maryland community can access Google Gemini by signing in with UMD credentials at gemini.google.com. Upon login, the central prompt window displays five icons: a Plus (+) icon for uploading documents, a Guided Learning icon for step-by-step topic exploration, a Deep Research icon for comprehensive answers, a Canvas icon for creating documents or mini-applications (this is not the same as ELMS-Canvas), and an Image icon that utilizes Google’s Imagen model to convert text descriptions into high-quality visuals. 

The available Gemini models include Flash for quick, everyday tasks and the more powerful Pro for complex reasoning and in-depth analysis. These models are frequently updated so keep an eye out for version changes!

Gemini integration with Google Workspace

Gemini integrates with key Google Workspace applications to extend functionality across common academic workflows. It can retrieve schedule information from Google Calendar, reference notes stored in Google Keep, and access task lists in Google Tasks. Drive files—Docs, Sheets, and Slides—are searchable within the chat interface, enabling summary and question-based interactions. It can search through Gmail to locate key information—such as specific messages, dates, or topics—based on natural language queries in the Gemini interface.

To connect Gemini with your Google Workspace, visit gemini.google.com, select Settings & Help at the bottom of the left sidebar menu.

Then, tap on Apps.

Scroll down to the Productivity section to enable access to apps such as Google Keep (keep.google.com) for virtual sticky-note reminders and Google Tasks (tasks.google.com) for to‑do list management. Once activated, queries like “What’s in my calendar for today” cause Gemini to search Google Calendar on your behalf. This Gemini integration can be toggled off at any time by using the slider switch located at the top right of the Productivity section.

NotebookLM powered by Gemini with audio and video capabilities

Other Google AI-powered apps include NotebookLM (notebooklm.google.com), which offers an AI-driven research and note-taking environment leveraging the Gemini models. Users can upload documents such as PDFs, Google Docs, or webpages to initiate conversational queries. The platform supports drafting research outlines, generating references, and refining documents within a single integrated workspace. One distinctive capability of NotebookLM is its audio and video summary features based on these uploaded documents, transforming content into engaging, concise overviews.

Under the hood, Gemini supports a one-million-token context window, equivalent to approximately 1,500 pages of text or 30,000 pages of text or thirty thousand lines of code. With around 100 tokens corresponding to 60 to 80 English words, substantial inputs can be processed without loss of context. Usage alerts notify users as daily limits are approached, ensuring continuity of service. Monitoring these notifications and optimizing token usage helps maintain Gemini’s effectiveness for diverse projects at UMD.

Limits on use of Google Gemini

There are limitations on how often you can use Google Workspace with Gemini. Specifically,

• Per-user monthly quota: Each individual with a Google Education license is allowed up to 1,000 uses of the Gemini app and its AI features per calendar month.
• Reset schedule: All users’ quotas automatically reset on the first day of each month.

These “uses” encompass any interaction with Gemini in Workspace—such as generating or summarizing content in Gmail, Docs, Slides, Sheets, or chatting with the Gemini sidebar—and once a user hits 1,000 interactions, they must wait until the next month for their allowance to renew.

The future of AI products from Google

Google continues to evolve its AI offerings and capabilities. New Gemini models or additional AI-powered features may become available over time, so UMD users are encouraged to check gemini.google.com for the latest updates and explore new tools as they are released.

For more information

If you have further questions about Google Gemini or other AI products from Google offered through Google Workspace, feel free to email the DIT AI Solutions Team at: dit-ais@umd.edu.