Live Webinar

AI Hot Topics: Breaking and Hardening the MCP Trust Chain

October 27, 2026 | 11:00AM ET / 4:00PM GMT

Your AI agent doesn't just write text anymore. It reads, decides, calls tools, and changes real systems, and MCP is the bridge that makes it possible. That shifts the security question from "is this API safe?" to "is this delegation safe?" Join Vandana Verma Sehgal for a look at how attackers are exploiting MCP in the wild, including the incidents and breaches making headlines right now, and the controls that stop an agent from becoming an attacker's easiest path into your environment.

What you'll learn:
  • How recent AI security incidents and breaches actually unfolded, and which MCP weaknesses they exploited
  • Where the MCP trust chain breaks: poisoned tool descriptions, confused deputy delegation, token and context leakage
  • How to apply the OWASP MCP Top 10 as a risk lens instead of a checklist
  • Practical hardening you can ship: server inventory, descriptor integrity, action-level authorization, sandboxing, and audit trails
Plus, live attendees will receive 1 CPE credit!

Save Your Spot

Who should attend?

  • Security leaders and AppSec teams who need to set policy for agents before they're already running in production
  • Platform and DevOps engineers standing up or maintaining MCP servers and agent integrations
  • Developers and AI engineers building with MCP who want to know where the trust boundaries actually sit
  • CISOs and security architects who need a defensible answer when the board asks how agent access is controlled

Date & Time

October 27, 2026 | 11:00AM ET / 4:00PM GMT

Speaker

Vandana Headshot 2026.png
Vandana Verma Sehgal
Staff Developer Advocate, Snyk

Snyk is trusted by the world’s most innovative companies.

© 2026 Snyk Limited
Registered in England and Wales
AISecEng Logo