Thursday, November 19, 2026

11:30 AM - 12:30 PM

Swan Room

---

Presentation Outline

Autonomous agents and AI-enabled decision-support systems are rapidly emerging across government and defence networks, from ISR processing to mission planning and logistics. As these systems move from experimentation to operational deployment, they introduce a new security perimeter defined by data flows, APIs and machine identities rather than human users alone. This tutorial provides a technology-focused introduction to securing AI-driven mission networks and autonomous agents in government and military contexts. We will explore how agents interact with sensors, data lakes, mission systems and external AI services, and examine key threats including model and data poisoning, prompt and command injection, supply-chain compromise, and rogue or misconfigured agents. Building on zero trust principles, the tutorial will introduce architectural patterns for strong identity, attestation, authorization and policy enforcement for non-human actors across enterprise, cloud and deployed environments. Participants will work through interactive scenarios in small groups, mapping attack paths and collaboratively designing controls such as secure enclaves, policy-as-code, segmented data pipelines and robust logging and assurance mechanisms. By the end of the session, attendees will understand the core security challenges of AI-enabled mission networks and leave with practical design patterns they can adapt to their own organisations.

---

Speaker