Cornell is expanding access to ChatGPT Edu, Claude Enterprise, Gemini, and Copilot along with other AI-related resources and services. These tools offer a range of capabilities, from content creation and data analysis to research support, coding assistance, and ways to streamline tasks.  

Broad access helps ensure that all members of the Cornell community have opportunities to examine how these technologies can impact their work, studies, teaching, and research. 

As part of this effort, Cornell has established pilots with providers to support a period of learning, experimentation, and evaluation across the university. Access is being introduced through a phased rollout, with different tools and user groups becoming eligible at different times throughout 2026 and early 2027. This approach allows Cornell to gather feedback, identify effective use cases, expand training and support resources, and incorporate lessons learned as use grows. 

Cornell has intentionally adopted a multi-vendor AI ecosystem to give the community meaningful choice. No single AI model or platform is best for every task, discipline, or use case. By providing access to multiple tools, Cornell enables individuals and teams to determine whether and when a tool will support their work, including in teaching, learning, and research. 

For the most current list of available AI tools, including details on eligibility, access, approved uses, training resources, privacy protections, and university guidance, please refer to Tools and Guidelines and Resources. 

For the Ithaca-supported campuses and Cornell Tech, access begins in November for faculty, staff, and Ph.D., DMA, and JSD students, with master’s students and undergraduates to follow in early 2027.   

Eligible faculty, students, and staff can request a base level of use for all tools at no cost, and faculty, staff, and teams will have the option to purchase additional capacity at a discounted rate. Free access to a limited number of licenses for Google’s agentic AI tools for research use will follow in spring. 

Because Weill Cornell Medicine operates in a distinct clinical, research, regulatory, privacy, and security environment, it will establish its own phased implementation plans, eligibility criteria, permitted uses, training requirements, and safeguards. Details are being finalized and will be communicated separately. Until then, members of the Weill Cornell Medicine community should continue to follow existing institutional policies and approved-tool guidance.

This phased approach allows Cornell to expand access thoughtfully while providing training and support, gathering community feedback, and incorporating lessons learned as use grows across the university. It also supports Cornell's ability to evaluate how AI tools can best support teaching, learning, research, and operations while informing future decisions about AI investments, governance, and support.  

For the latest information on availability, eligibility, and rollout timelines, visit the Tools and Guidelines and Resources pages. 

Yes. Cornell offers a growing range of training, learning opportunities, and support resources to help faculty, staff, and students build AI skills and use AI tools effectively and responsibly.   

Resources include the AI Critical Literacy Program, the AI Exploration Series, tool-specific training, selected eCornell offerings, and educational materials from trusted learning platforms and AI providers. Community members can also connect with peers, exchange ideas, and share lessons learned through communities of practice, Teams channels, and other collaborative forums. 

Coming soon: Cornell will expand its offerings with additional workshops, demonstrations, office hours, and hands-on training sessions, including tool-specific deep dives and role-based learning opportunities. 

Visit the Cornell AI Initiative Training page for the latest resources, events, and support opportunities. 

Cornell is taking a phased approach to expanding access to AI tools across the university. This approach allows the university to thoughtfully scale services, incorporate lessons learned, develop training and support resources, and ensure users have a positive experience as new tools are introduced. 

Faculty, labs, and academic units that would benefit from earlier student access for teaching, research, or other university activities may have opportunities to sponsor access for students within their courses, programs, labs, or project teams, subject to tool availability and program requirements. Submit an email to start the request.

Faculty, students, and staff are already bringing AI tools purchased on consumer licenses to campus. These licenses have terms that create substantial risks to privacy, security, loss of intellectual property, and security. 

By providing university-supported AI tools, Cornell has been able to negotiate institutional pricing, privacy and security safeguards, and contractual protections that are not available through individual consumer subscriptions. Institution-wide access also helps promote equitable participation by ensuring that advanced AI capabilities are available to the broader Cornell community, rather than only to those who can afford individual subscriptions. It also helps ensure that all faculty, staff, and students, regardless of role, discipline, or technical background, have the opportunity to explore how these technologies can enhance their teaching, learning, research, and work, while engaging with AI in ways that are secure, responsible, and aligned with university values. Through individual experimentation, community-wide training opportunities, and novel application and iteration, our campus community can gain deeper understanding of how to mitigate AI risks and how to take advantage of new opportunities.

Use of Cornell-supported AI tools is generally optional, and community members may decide whether and how AI is appropriate for their teaching, learning, research, or work. Students should talk to their instructor about whether classroom use of AI is expected.

Cornell's current AI tool pilots are designed to support learning, experimentation, and evaluation across the university. During this period, faculty, staff, and students are encouraged to explore AI tools where they may be helpful, while also sharing feedback about when AI adds value, when it does not, and where limitations or concerns exist.

Cornell recognizes that AI is not the right solution for every task, discipline, workflow, or learning objective. Decisions about whether and how to use AI should be guided by professional judgment, disciplinary norms, applicable policies, and individual needs.

Students should follow guidance provided by their instructors and any course-specific policies outlined in the syllabus. Staff should consult with their supervisors and unit leadership regarding appropriate use within their roles and responsibilities. Faculty should consider disciplinary norms, applicable university policies, and any relevant research, teaching, or professional requirements.

Feedback from the community, including both successful use cases and situations where AI is not useful or appropriate, will help inform future decisions about AI tools, training, guidance, and investments.

Like other Cornell-provided computing, networking, and information resources, AI tools are provided primarily to support Cornell's teaching, learning, research, and workplace activities. In most cases, incidental personal use is permitted, provided it is reasonable, does not add costs to the institution, does not interfere with university work or studies, and complies with applicable laws, university policies, and contractual obligations (see Policies and Guidelines for more information).

Users are expected to exercise good judgment, act responsibly, respect the rights of others, and use these resources in accordance with university policies and standards of conduct. Users should not use university-provided AI tools in ways that create legal, security, privacy, or reputational risks for Cornell, interfere with university operations, or violate university policies, vendor terms of service, or applicable laws.

Use of university IT resources that results in inappropriate workplace conduct may be addressed as an employee performance issue, even if the conduct does not rise to the level of a university policy violation. Any use of university computers, information resources, and networks by employees that is inappropriate to the workplace, contributes to a harassing or uncomfortable work environment, or creates a legal risk, will result in counseling, formal disciplinary action, and/or termination. Such performance concerns should be directed to a supervisor or the unit Human Resources representative.

If you created a personal Claude account with your NetID@cornell.edu email address, you can switch between your Cornell-provided and personal accounts.

If you created a ChatGPT account with your NetID@cornell.edu email address, you will be required to merge your personal account with the Cornell-provided account. This is not always appropriate or desired. Before you complete the AI Training course, review this important information about how to prevent this conflict.

If you merge a personal ChatGPT account to your Cornell account, this action cannot be undone.

Just like any other login service (such as CUWebLogin and Microsoft Entra) you may only use your NetID@cornell.edu email address to log in to your Cornell-provided Claude and ChatGPT accounts. Email aliases cannot be used for authentication to these services.

In general, AI tools do not support bulk migration of chats, projects, or conversation histories between accounts. Some providers offer account migration options that may transfer certain content and settings, but these processes are not always comprehensive and may not transfer all data, files, customizations, or project history.

For tool-specific guidance, refer to Anthropic’s account migration information for Claude, and OpenAI’s migration guidance for ChatGPT.

If you created a personal Claude account with your NetID@cornell.edu email address, you can switch between the Cornell-provided and personal accounts.

If you maintain separate ChatGPT accounts, you must sign out of one account before signing in to the other. ChatGPT does not currently support switching between separate accounts within the same session.

When working with Cornell data, always use your Cornell-supported account rather than a personal account. Cornell-supported accounts include privacy, security, and contractual protections that may not be available through personal subscriptions.

Cornell's AI policies, guidance, and recommended practices are available on the Resources page.

The page includes resources on responsible AI use, privacy and data protection, information security, accessibility, research considerations, and other university policies that may apply when using AI tools. Faculty, staff, and students are encouraged to review these resources before using AI for teaching, learning, research, or operations.

Because AI technologies, capabilities, and guidance continue to evolve, Cornell regularly updates its resources. Community members should refer to the site periodically for the latest policies, recommendations, and best practices.

For tool-specific guidance, including approved uses, data handling requirements, and privacy considerations, see the Cornell-Supported AI Tools secion on this page.

Cornell has negotiated agreements that include enhanced privacy and data protection safeguards, including protections against using your Cornell data to train the providers' models. Cornell-supported AI tools are designed to provide stronger privacy and security protections than public or consumer AI tools. These protections support the use of AI tools for teaching, learning, research, and administrative work while helping ensure alignment with Cornell's data management and security requirements.

However, privacy and security protections do not eliminate the need for good judgment. Users should to follow Cornell policies and data handling requirements when using AI tools. While Cornell-supported tools include additional protections, users remain responsible for handling university information appropriately and should follow university policies and requirements when working with confidential, regulated, proprietary, or otherwise sensitive data. 

Non-Cornell or public AI tools are different. Even if you are paying a subscription to a public tool the data is still at risk. Unless specifically reviewed and approved by Cornell, information entered into publicly available AI services may be stored, retained, or used by the provider in ways that differ from university-supported tools. Do not enter confidential, proprietary, regulated, restricted, or another person's sensitive information into public AI tools.  

If you are unsure whether a tool is approved for university use or whether specific data may be entered into a tool, consult Cornell's AI guidance and data handling policies before proceeding. 

The types of university data that may be entered into AI tools depend on the specific tool, the data involved, and the protections available within that service. Before using an AI tool, users should understand both the tool's capabilities and the university's requirements for handling institutional data.

All Cornell-supported AI tools prohibit the use of high-risk data, and some data types require additional precautions. For example, when working with FERPA-protected data or international personal data in Microsoft Copilot or ChatGPT, web search capabilities must be disabled before that information is entered into the tool (learn how to do this for ChatGPT).

In general, Cornell-supported AI tools provide stronger privacy and security protections than a public or consumer AI tools and may be appropriate for a broader range of university activities. However, users remain responsible for following applicable university policies, data classification requirements, contractual obligations, sponsor requirements, and legal or regulatory restrictions when using any AI tool.

Refer to this page for specific guidance on what types of university data may be used with each AI tool, including any restrictions, approvals, or special requirements.

If you accidentally enter sensitive, confidential, regulated, or otherwise restricted information into an AI tool, delete the conversation or content immediately, if the platform allows it.

You should then report the incident to the IT Security Office by sending an email to itsecurity@cornell.edu, just as you would report any other potential data exposure or information security incident. Prompt reporting helps the university assess potential risks, determine whether additional actions are needed, and provide appropriate guidance.

If you are uncertain whether the information was sensitive or whether the incident needs to be reported, err on the side of caution and contact the IT Security Office for assistance.

Reporting an accidental disclosure promptly is an important part of protecting university data and supporting responsible AI use.