Speaking

Speaking & engagements

Sessions for engineers and security leaders on agent attacks, runtime controls, and practical cybersecurity architecture.

Accepted · October 13, 2026 · Online

ConfusedPilot: Breaking Delegated Authority Across AI Agents and Tools

[re]aligned, the online extension of CanSecWest. An accepted talk on tracing, limiting, and revoking delegated authority.

Event ↗

October 9, 2026 · Richardson, Texas

Breaking AI: Real-World Attacks & Defenses for Agentic AI Systems

An interactive University of Texas at Dallas session on agent attacks and runtime defenses.

Official event ↗

July 2026

Securing AI Agents: Guardrails and Runtime Controls

ISC2 Think Tank speaker and panelist, covering secure autonomous enterprise workflows.

Topics

Talks grounded in real workflows.

Agent identity and authorization · Prompt injection and tool abuse · AI incident response · Threat modeling autonomous systems · Application, API, cloud, and DevSecOps security.

Discuss a session ↗