Vancouver Meetup #1

Joining us from Durban, South Africa — Ms Janet Bruce-Brand, Lecturer at the University of KwaZulu-Natal, delivers the featured keynote: *"Ubuntu for Responsible AI: Leadership Lessons from Higher Education and the Global South."*

Session description

AI is now embedded in how institutions teach, assess, research and decide. The governance built around it is mostly compliance — policies written to manage institutional risk rather than to serve the people the systems act on.

Janet Bruce-Brand argues for a different starting point. Ubuntu — *"I am because we are"* — treats accountability as relational rather than procedural, and it produces different answers about who has standing in a decision, what counts as harm, and which trade-offs an institution may accept on someone else's behalf.

Drawing on her work in a university AI Thought Leadership Forum developing institutional guidelines, and on doctoral research into ethical integration, academic integrity, governance and capacity-building, she examines what responsible AI actually requires where digital inequality, constrained resources, language diversity and historical educational inequity are not edge cases but operating conditions.

The lessons travel. Any organisation deploying AI across communities it does not resemble faces the same question: whether its governance was designed for the auditor or for the people it affects.

Key takeaways for senior delegates

1. Governance from the Global South, not about it. What happened when African universities met generative AI — which policy drafts failed first, where institutional readiness actually broke, and what academic integrity work survived contact with students. 2. Ubuntu as a working framework, not a metaphor. Relational accountability applied: how harm is defined when it lands on a community rather than an individual, and what changes when dignity is treated as a design constraint rather than a values statement. 3. How institutional AI guidelines get built. Transferable practice on engaging stakeholders who did not ask for the technology, supporting AI literacy across uneven skill levels, and holding innovation and safeguards in the same policy. 4. Constraint as a design input. Unequal infrastructure, under-represented languages and local knowledge systems shape what can responsibly be deployed — a discipline that matters anywhere a system serves populations its designers did not model. 5. Alignment is the hard part. Responsible integration needs governance, teaching practice, assessment design, staff development and student support moving together. The common failure is a sound policy nobody implements.