GDG Student Club — Featured Lecture: Security in AI Agents

As AI agents move beyond chatbots to systems capable of autonomously accessing data, calling tools, and taking actions, they introduce an entirely new set of security challenges. In this guest lecture, Khushan Adatiya, author of the 2026 book Securing the Agentic Enterprise: Threat Modeling, Anomaly Detection, and Governing Autonomous Multi-Agent Systems, will explore security for LLM-based and agentic AI systems. The talk will examine vulnerabilities and attack vectors unique to these systems. Drawing on real-world systems and industry practices, the session will give students a practical introduction to designing AI agents that can operate securely in production environments.

About the Speaker: Khushan Adatiya

Khushan Adatiya is a Senior Software Engineer at Google whose work spans large-scale distributed systems, AI infrastructure, API security, and AI-driven observability. He has more than a decade of experience building large-scale software systems, including engineering roles at both Google and Amazon Web Services. His recent work focuses on agentic AI, cloud architecture, security, anomaly detection, and the infrastructure needed to deploy AI systems reliably at scale. He has also presented work on AI agents and cloud infrastructure at venues including the IEEE Cloud Summit and DeveloperWeek/AI DevSummit. Khushan is the author of Securing the Agentic Enterprise and has published on topics including AI-agent security, multi-agent systems, scalable AI infrastructure, and agent-to-agent communication.