
November 23–27, 2026
Senior Manager
EXLService
AI agents are quickly moving beyond answering questions. They can retrieve enterprise data, call tools, interact with internal systems, and increasingly take actions on a user's behalf. But as agents become more capable, an important question emerges: just because an agent can access a piece of data, should it? This talk looks at agentic AI through the lens of enterprise data governance. Drawing from my experience working with complex data flows in regulated financial environments, I will discuss what changes when AI agents become consumers of enterprise data. We will look at five practical questions: What data does the agent actually need? Can we trace where that data came from? Is the data reliable enough for the task? What actions should require additional controls or human review? And who is accountable when something goes wrong? Attendees will leave with a practical framework built around Know, Limit, Trace, Control, and Own that they can use when thinking about data access for enterprise AI agents. Rather than treating governance as something added after an agent is built, the session will show how data lineage, access boundaries, data quality, traceability, ownership, and human checkpoints can become part of designing trustworthy agentic systems from the beginning.
See every session across the AI Agents Conference November 2026 line-up.
View Agenda