New AI Tools for Stanford Arrive June 30

OpenAI ChatGPT Edu, Google Gemini Enterprise, and Anthropic Claude for Education will be available to Stanford faculty, students, postdocs, and staff on June 30. Details on how to get access to these tools from University IT (UIT) are below.

These offerings are part of a campus pilot through August 2027. Depending on your role at Stanford, additional approval is required before using plugins, connectors, or automated workflows with these new AI tools.

For questions about the AI tools approved for use at Stanford, visit our new AI Services Matrix. Learn more about AI training and responsible AI use on our UIT AI page.

How do I get access?

OpenAI ChatGPT Edu

On June 30, all Stanford faculty, students, postdocs, and staff will have access to ChatGPT Edu, giving the Stanford community access to powerful AI tools for academic and administrative work.

ChatGPT Edu access requires submitting a simple request form on the service page (under Get Started). You’ll need to confirm your Stanford identity through single sign-on and agree to the responsible use terms before your account is activated.

• ChatGPT Edu supports writing, research, and data analysis using the highly accurate GPT-5 model.
• Codex can work across your files to help you build custom apps, websites, presentations, and workflows.
• Open API connects with OpenAI models, research tools, or prototypes to generate text, voice, and audio.

Google Gemini Enterprise

All Stanford faculty, students, postdocs, and staff with a Stanford Google account will have access to Gemini Enterprise starting June 30. This tool activates automatically the first time you sign in with your SUNet ID through the Gemini Enterprise homepage (under Get Started).

Gemini Enterprise Standard gives you access to Google's latest AI models within the productivity tools you already use:

• NotebookLM Enterprise allows you to upload your own sources, which is helpful for literature reviews, course preparation, and grant writing.
• Gemini Enterprise can connect to Google Drive, OneDrive, Slack, and Microsoft 365 (e.g., Outlook and Teams) for all university users.
• Agent Designer allows users to create, manage, and launch AI agents.

Anthropic Claude for Education

Like ChatGPT Edu, Claude for Education access requires submitting a simple request form on the service page (under Get Started). You’ll need to confirm your Stanford identity through single sign-on and agree to the responsible use terms before your account is activated.

Claude for Education includes three tools, each suited to different kinds of work:

• Claude Chat is a powerful conversational AI for writing, research, analysis, and document review.
• Claude Code is an AI coding assistant that runs in your terminal and is well suited for research software development and complex programming tasks.
• Claude Cowork is a desktop agent that automates multistep tasks across your files, calendar, and connected apps, including Google Drive, Slack, Zoom, and Microsoft 365.
  • For School of Medicine users: To protect patient privacy (HIPAA), Zoom, Outlook, and Teams connectors are turned off.

Data security and compliance considerations

As the Stanford community gains access to these powerful tools, please note that connectors, plugins, and repeatable workflows (such as Claude Skills) will require additional checkpoints before they are approved for use.

These checkpoints are intended to protect sensitive data (e.g., student records, financial transactions, and HR information) from unauthorized access or inclusion in systems of record, ensuring our core institutional data remains secure and compliant.

OpenAI ChatGPT Edu

ChatGPT Edu is approved for low-, moderate-, and high-risk data, as well as protected health information (PHI). This covers the vast majority of everyday academic, research, and administrative work.

Your conversations are not used to train OpenAI’s models and include additional copyright protections.

Google Gemini Enterprise

Google Gemini Enterprise is approved for low-, moderate-, and high-risk data, as well as protected health information (PHI). This covers the vast majority of everyday academic, research, and administrative work. Your conversations are not used to train Google's models, and the same data-handling commitments that apply to Stanford Google Workspace apply here.

Anthropic Claude for Education

Claude for Education is available to all faculty, students, postdocs, and staff. Please note these data risk classifications before you begin your work with these tools:

• Claude Chat is the recommended starting point for most users and is approved for all data classifications except protected health information (PHI).
• Claude Code is approved for low-, moderate-, and high-risk data. PHI use requires routing through the Stanford AI API Gateway.
• Claude Cowork and its connectors (Google Drive, Slack, Zoom, and Microsoft 365) are approved for low-, moderate-, and high-risk data. PHI use requires routing through the Stanford AI API Gateway.
• For School of Medicine users: Due to HIPAA compliance considerations, Zoom and Microsoft 365 (e.g., Outlook and Teams) connectors are disabled regardless of data classification.
• If you are using Claude for federally supported research, there may be additional restrictions to consider. Please contact the Information Security Office if you have questions.

If you already have a personal Claude account associated with your Stanford SUNet ID email address, you will need to migrate to the new Stanford Claude Enterprise workspace by October 15.

Need more? Upgrade to ChatGPT Pro, Gemini Enterprise Pro, or Claude for Education Premium

For users with heavier research, productivity, or workflow needs, ChatGPT Pro, Gemini Enterprise Pro, and Claude for Education Premium are available at discounted rates negotiated exclusively for Stanford.

Monthly charges will be prorated from your activation date through the end of the contract period for each service. These licenses are charged to a Stanford PTA and require approval from a PTA approver.