This session addresses the next big trend in AI innovation: AI agents. Agents remain a nascent technology, and the opportunities and risks they present are not well understood. Cross-community engagement is needed to develop greater understanding about what a responsible approach to building and deploying AI agents looks like, and how best to ensure user safety, privacy and agency in this new context.
The discussion will then consider how to embed user-centricity in the building and deploying of agentic products. How should companies govern what data agents should have access to? How do we manage the trade-off between privacy and effectiveness? How can we balance agentic autonomy with rigorous oversight and end-user control?