The University of California, San Francisco (UCSF) is rolling out OpenAI’s ChatGPT Enterprise platform to its community, replacing its existing Versa Chat system. The move, announced by UCSF Senior Vice President and Associate Vice Chancellor Joe Bengfort, aims to provide secure access to advanced AI capabilities for research, education, clinical care, and administrative functions.
ChatGPT Enterprise builds upon UCSF’s prior work with its Versa platform, which offered private and secure access to generative AI. The new platform leverages OpenAI’s latest models, including GPT-5, and incorporates enterprise-grade safeguards, notably enhanced support for compliance with the Health Insurance Portability and Accountability Act (HIPAA). This is critical for an institution handling sensitive patient data.
The rollout will occur in two phases. Existing Versa Chat users – approximately 9,000 individuals – will begin gaining access on . Access will then be opened to all other UCSF community members with ucsf.edu accounts starting on , with users required to request access.
UCSF’s decision to adopt ChatGPT Enterprise isn’t simply about upgrading to a newer model. It’s a strategic shift towards a more robust and secure AI infrastructure. The university has been actively exploring generative AI for some time, and the Versa platform demonstrated the value of providing secure access to these technologies. However, ChatGPT Enterprise offers a more comprehensive solution with features designed for large organizations, including enhanced data encryption, privacy controls, and enterprise-level security measures.
According to Bengfort, ChatGPT Enterprise includes safeguards such as HIPAA compliance and protection for data classified as P3/P4 – UCSF’s highest levels of data sensitivity, typically involving dangerous pathogens. This level of security is paramount for maintaining the integrity of research and protecting patient privacy.
The transition won’t be immediate for all users. While the 9,000 Versa Chat users will be migrated first, the university also operates over 100 “Versa Assistants” – likely automated workflows or chatbots built on the Versa platform. The migration of these assistants will occur in a subsequent phase, with a timeline to be announced soon. UCSF will also be providing OpenAI training opportunities to help the community effectively utilize the new platform.
Importantly, UCSF’s Versa API will continue to operate, providing secure API access to large language models from multiple vendors, including AWS, Azure, Anthropic, and OpenAI. This allows developers within UCSF to continue building custom AI applications using a variety of models and tools. Existing Versa Chat users will retain access to Versa Chat and their Versa Assistants until further notice, ensuring a smooth transition.
The move to ChatGPT Enterprise reflects a broader trend in higher education and healthcare towards adopting secure, enterprise-grade AI solutions. Universities and hospitals are increasingly recognizing the potential of generative AI to accelerate research, improve patient care, and streamline administrative processes. However, they are also acutely aware of the risks associated with data privacy and security. ChatGPT Enterprise, with its focus on data protection and HIPAA compliance, addresses these concerns directly.
UCSF’s implementation of ChatGPT Enterprise is not just a technology upgrade; it’s a statement about the university’s commitment to responsible AI innovation. By prioritizing security and privacy, UCSF aims to unlock the full potential of generative AI while safeguarding its data and maintaining the trust of its community. The university’s experience with Versa Chat provided a foundation for this transition, and the adoption of ChatGPT Enterprise represents a significant step forward in its AI strategy.
Further details, including FAQs, are available on the UCSF AI website: https://ai.ucsf.edu/ucsf-chatgpt-enterprise-faqs.
