In November 2025, Anthropic disclosed what it describes as the first documented cyber-espionage campaign orchestrated by an autonomous AI system. This presentation will unpack how advanced AI agents transformed reconnaissance, exploitation and exfiltration from human-driven processes into machine-driven operations, explore why conventional detection and governance frameworks are now inadequate, and map out practical strategies for designing assurance, telemetry and governance regimes aligned with this new paradigm. Attendees will gain insight into how trust, oversight and defence must evolve when adversaries deploy agentic AI systems—and what mission-assurance professionals, regulators and security operations teams need to consider to maintain resilience.
