AI is evolving into autonomous agents that connect with multiple business systems via the Model Context Protocol (MCP). While these agents boost efficiency, they also introduce new security risks. They can access sensitive data and act across systems, sometimes without human oversight. This session covers key risks, real-world examples of failures, and practical ways to keep AI agents secure and trustworthy.
talk-data.com
X
Speaker
Xia Hua
1
talks
CEO
Traceforce
Xia Hua is CEO of Traceforce, where she builds security controls to help enterprises govern and secure AI applications. She previously led engineering at Clumio (acquired by Commvault), delivering cloud data protection products that were 20x faster and 10x more scalable than competitors. Earlier, she spent nearly a decade at Oracle contributing to the 12c In-Memory Database. Xia earned her PhD in Applied Mathematics from MIT.
Bio from: Securing the Next Frontier: Understanding Risks and Protections in Agentic AI
Filter by Event / Source
Talks & appearances
1 activities · Newest first