Berlin Cybersecurity Social #25

​Securing the AI Frontier: Navigating the Security Risks of the Model Context Protocol (MCP)

​As organizations move from static API integrations to dynamic, AI-driven agent systems, new security risks are emerging that traditional security models were never designed to handle. This month's Berlin Cybersecurity Social explores how protocols like the Model Context Protocol (MCP) reshapes the threat landscape by enabling agents to dynamically access tools, data, and permissions.

​Siva will break down the unique risks introduced by MCP-style architectures, including identity sprawl, uncontrolled access, and limited visibility into agent behavior. The session will also cover practical, actionable defense strategies to help security teams adapt their threat models, controls, and governance for an agentic AI world.

​

​About Siva Rajendran

​
Siva Rajendran (OSCP, CRTE, CEH) is a Senior Cybersecurity Auditor at Tesla. He is an active security researcher on platforms such as Synack, Intigriti, and Bugcrowd, where he has earned multiple awards for his work. His background spans offensive security, auditing, and real-world vulnerability research, bringing a hands-on perspective to emerging AI security challenges.