Marcus Smolka, professor of molecular biology and genetics, was impressed. He’d fed an AI tool a complex dataset, typical in his research, that analyzed thousands of proteins. Not only could the tool glean insights from the data, it also suggested – correctly – another dataset as a control for the experiment. 

“I was like, ‘Wow,’” said Smolka, who gained early access to a Cornell-supported Claude Code account, one of a suite of tools – including OpenAI ChatGPT and Codex, Anthropic Claude Cowork and Code, Microsoft CoPilot Chat, and, tentatively, Google Gemini Chat – that the university will make available, at no cost, to all eligible faculty, staff and students in the coming months.  

“These tools are going to be very important to stay competitive,” Smolka said, “and we need to bring people together to build systems checks and balances and to find best practices and workflows.” 

Base-level accounts will be available to faculty, staff and doctoral students on the Ithaca campus and at Cornell Tech in November, with access for master’s students and undergraduates coming in early 2027. Free access to a limited number of licenses for Google’s agentic AI tools for research will follow in spring. Weill Cornell Medicine is planning a phased release of Claude in the coming weeks for use cases that don’t involve protected health information.  

With expanded access across campuses, Cornell hopes to level the playing field and spur exploration and innovation, as well as reduce the costs and vulnerabilities of individual consumer licenses. The new contracts will ensure enterprise-level protection for security and privacy – data entered into the new tools cannot be used for training new AI models – and the university will be able to better track energy use related to AI. 

The effort aligns with the recent Future of the American University report, which calls for a human-centered approach to AI, said Provost Kavita Bala. 

“AI is here, and it is influencing many core aspects of the university’s mission,” Bala said. “We want to position Cornell as a leader in the dialogue that develops ethical and human-centered AI – and show how best to take advantage of the immense opportunities and meet the considerable challenges head-on. This is not a technology project but a commitment to the long-term strength of Cornell’s teaching, research, operations and public service.” 

Thorsten Joachims, vice provost for AI strategy and director of the Cornell AI Initiative, said that the work ahead will often include deciding when to intentionally not use AI. 

“We need to preserve and nurture human judgment, which keeps us centered on the university’s mission,” he said. “Our scholarship on ethical and human-centered AI is essential, because we have an important responsibility to add to the dialogue in a way that industry cannot. AI is not going away, and we need to have a seat at the table.” 

Meeting the opportunities and the challenges 

Joachims coordinates with leaders across campus – part of an AI Strategy Council – to support scholarship around AI as well as community-based experimentation on how to appropriately integrate the technology into teaching, research, operations and public engagement. 

“We made sure to have programs in place that support the community in making responsible use of these AI tools,” he said. He mentioned the AI Critical Literacy Training for undergraduate students, the new research integrity policy for AI, the Communities for Practice for AI-enabled Research (CoPAIR), and the AI@Work program in the AI Innovation Hub for staff.  

“And getting access to agentic tools will include a primer on responsible and appropriate use,” Joachims said.  

The platforms can help with tasks ranging from content creation and data analysis to research support, coding and the streamlining of administrative work. 

Providing access to tools from multiple vendors allows users to determine the best option for their specific tasks and also gives the university flexibility in the future, said Ben Maddox, Cornell’s chief information officer and a leader of the AI effort. Access will run for a two-year pilot period, at which point the university can assess use of the tools as well as the external AI landscape. Sustainable Computing, a working group of faculty, staff and students, will track AI’s impact on energy use and make recommendations as well. Users will have access to basic accounts, and some will be able to purchase additional capacity at a discounted rate.  

The AI Innovation Hub, led by Ayham Boucher, executive director of AI strategy, innovation and enablement, is coordinating the tool rollout, technical support and training opportunities for faculty and staff. A required orientation will prime faculty and staff for responsible AI use, and additional training will be available through the AI Innovation Hub, from the vendors, from eCornell and from third parties such as Workday Learning. 

“We want our community to have firsthand knowledge of the opportunities and challenges of AI tools,” Boucher said, “And to bring that knowledge back in the form of best practices and recommendations about how and in what ways AI might or might not impact our collective work.” 

Incoming students already complete a required AI critical literacy training, and many will encounter and learn to use AI in their courses, as faculty find innovative ways to integrate its use, said Steven Jackson, vice provost for academic innovation.    

Beyond training, multiple programs within colleges and units will engage faculty and staff in the specific context of their work. The CoPAIR program, for example, will bring researchers together to develop norms and best practices for their disciplines.  

Smolka, who is also an associate vice provost in Cornell Research and Innovation and interim director of the Cornell Institute of Biotechnology, is coordinating the CoPAIR pilot for biotechnology researchers, who often work with enormous datasets that AI can help organize and analyze, speeding and facilitating breakthroughs in understanding biological systems. He hopes faculty, graduate students and staff can explore as a group and share what they’ve learned; more than 100 people have already registered. He said concerns over data privacy have kept many researchers from fully embracing the tools to date. 

“It’s often the first question I hear from faculty,” he said. “The privacy protection with the new tools is essential for a lot of researchers here who use confidential or sensitive data, so this opens a lot of doors for people to be more comfortable.” 

The data protection is also key for administrative staff. Ingrid Jensen, director of student services in Cornell Bowers, helped organize a training program for Bowers staff, “We Do AI,” last summer, which granted staff early access to the Cornell-supported tools. For the first time, her unit can now analyze troves of data, such as which courses appeal to students in certain majors, and the staff hours allocated to courses. 

“There are a lot of opportunities for staff who have data but not the time to get to it, especially when it comes to manually coding qualitative data,” she said. “And you have to be able to analyze the data to make data-driven decisions – so there are just real improvements we can make with these tools.” 

Jensen said the tools can also take tedious tasks off people’s plates. “If AI can help us spend less time on administrative tasks and more time engaging with students, solving problems and improving programs, that’s a meaningful benefit,” she said. 

Both Smolka and Jensen said that experimenting with the new tools impressed them but also reinforced the importance of human oversight. 

“A lot of people are realizing the incredible power of AI, but it doesn’t really replace the core aspects, like knowing how to ask a good question in biology,” Smolka said. 

Jensen said the more she engages with AI, the more she sees how crucial the human role is. “Critical thinking does not become less important in an AI-enabled workforce,” she said. “It becomes more important, because the output you’re getting is only so good as what you put in.” 

For more information about accessing the tools, visit the Cornell AI Initiative’s tools page.