As artificial intelligence rapidly reshapes operations and administration, the UC AI Council launched a four-part “AI Essentials” webinar series in spring 2025 to support faculty, staff, and administrators across the UC system. These sessions delivered practical guidance, real-world case studies, and actionable methods and discussed tools to help UC stakeholders responsibly adopt and manage AI technologies. If you missed them, all recordings are now available, and key takeaways from each session are summarized below.
Click on the image to see the video. The 1st session is only available to UC employees.
1. AI & Third-Party Risk: Best Practices for Evaluating and Procuring AI Solutions
Recorded March 26, 2025
[Watch Now (60 minutes) (UC Login Required)]
This session addressed the steps for responsibly evaluating AI tools, from pre-adoption risk assessment to long-term oversight. Panelists emphasized the importance of collaboration across procurement, IT, privacy and legal functions. Thank you, panelists: Lisa Ho, Bethanie Brown, Sajjad Matin, Gigi Capuyan
Top Takeaways:
- Define purpose first: Ensure AI is a sustainable fit for the problem you’re solving.
- Assess risk responsibly: Use UC’s Responsible AI Principles, prioritizing transparency, privacy and fairness.
- Protect privacy & accessibility: Scrutinize vendor practices and design for all users.
- Secure data ownership: Contracts should clearly define UC’s rights and limit liabilities.
- Plan for oversight: Build in regular reassessment and user training from the start.
Target audiences: Procurement teams, legal reviewers, IT leaders, and anyone evaluating AI vendors.
2. AI in Action: UC AI Innovation Virtual Share Fair
Recorded April 9, 2025
[Watch Now on YouTube]
This fast-paced session spotlighted three AI tools already in use at UC campuses, each designed to improve operational efficiency while upholding ethical standards. Thank you, panelists: Jennifer Lofthus, Diana Cox, Adam Getchell, Stefan Tomic, Anna Ahearn, Krithik Udayashankar
Featured Projects:
- Pose AI (Risk & Safety Solutions): Automates ergonomic assessments, supporting hybrid workforces.
- Policy Wonk (UC Davis): Answers complex UC policy questions using AI-powered search.
- Access Aid (UCLA): A chatbot that simplifies account provisioning and improves the IT support experience.
The session also introduced the UC AI Label, a transparency framework akin to a nutrition label, outlining an AI tool’s purpose, data handling, and risks.
Target audiences: Campus innovators, IT staff, and business process leaders exploring scalable AI solutions.
3. Generative AI and the Art of the Possible: Upskilling for the Future
Recorded May 21, 2025
[Watch Now on YouTube]
In this dynamic session, UC faculty from UCLA, UC San Diego, UC Irvine and UC Berkeley shared how they are using generative AI to transform teaching, learning and engagement. Thank you, panelists: Tina Austin, Kristina Markman, Sesh Murthy, Tamara Tate, and Erin Van Dusen and moderator Rolin Moe.
Highlights:
- Custom GPTs and PapyrusAI are enabling personalized learning and real-time feedback.
- Educators are rethinking traditional essays, leveraging AI to support writing scaffolds and Teaching Assistant (TA) workflows.
- Panelists emphasized AI literacy, ethical use, and human-AI collaboration as core teaching practices.
Instead of banning AI due to concerns over academic integrity, panelists advocated for transparency, thoughtful integration and student training.
Target audiences: Instructors, academic support staff and curriculum designers.
4. Safe Handling of UC Data in Generative AI Applications
Recorded May 29, 2025
[Watch Now on YouTube]
This webinar focused on how to protect sensitive institutional data while using generative AI tools like ChatGPT, Copilot, and Gemini. Thank you, panelists: Chris Mattmann (UCLA), Al Lavassani (UC Office of the President), Zainab Shakoor (UC Davis), Sesh Murthy (UC San Diego), Shea Lovan (UC Santa Barbara).
Key Takeaways:
- Know your data (P1–P4 signify levels of sensitivity/cost should they fall into the wrong hands): Data classification guides what can be shared with AI tools.
- Choose controlled environments: Prefer licensed tools over open web platforms.
- Collaborate with experts: Work with IT leads, privacy officers and data stewards.
- Be cautious with enterprise AI: Even tools within Microsoft 365 or Google Workspace need review.
- Emphasize “shared responsibility”: Ongoing education and clear policies are essential.
Target audience: Anyone working with protected or sensitive UC data—including administrative, IT, and research staff.
Final Thoughts
Across all four sessions, a common message emerged: AI should be used with care and transparency on the one hand, and collaboratively and proactively on the other. Whether you are exploring custom tools or evaluating vendor solutions, these webinars offer the insights needed to make informed, responsible decisions.
Questions?
Reach out to the UC AI Council at aicouncil@ucop.edu
