AI, MCP and Security

Everyone talks about AI. And, of course, its usage should be secure. But how possible - or impossible - is it to achieve this goal? In this presentation, we will discuss the current state of the art – particularly in the context of MCP and agentic AI systems, from attack vectors to IAM with Keycloak.

If you are asking yourself: „What is the threat situation? What options are there for authentication, authorization and runtime security in general? Is OAuth really the answer to everything? And how certain is it that everything will stay that way?” – then you’ve come to the right place. Several demos will give you an overview of attacks on AI and MCP, and ways to protect your MCP servers against them.

Datum

05.03.2026

Uhrzeit

11:00 - 11:45

Konferenz / Veranstaltung

Keycloak DevDay

Ort

greet Hotel Darmstadt, Darmstadt