From Autonomous Agents to Trusted Systems: Rethinking AI Agents in the Age of Automation, How Autonomous Al Is Redefining Cybersecurity Paradigms Control, and Resilience in Autonomous Systems. Secure and Trustworthy Machine Learning and AI for Multi-Domain Applications

The rapid evolution of artificial intelligence is shifting us from static, rule-based systems to agentic architecturesautonomous entities capable of reasoning, decision-making, and independent action. This transition is fundamentally redefining the cybersecurity landscape. In a world where AI systems are no longer just tools but active participants, the question is no longer what can be built, but what can be trusted.

This talk explores the emergence of agentic security, a new paradigm designed to secure ecosystems of intelligent agents operating at scale. We examine how traditional security models—built for deterministic systems—fail in environments where agents dynamically collaborate, evolve, and potentially act adversarially.

Through real-world architectures and forward-looking frameworks, the session introduces key pillars of agentic security, including agent identity and intent verification, continuous observability, trust orchestration, and self-healing defense systems. It also addresses critical challenges such as the role of humans in the loop, governance of autonomous decisions, and the risks of AI-vs-AI attack scenarios.

Attendees will gain a practical and strategic understanding of how to design, secure, and scale agent-driven systems moving from reactive defense to adaptive, intelligent, and resilient cybersecurity ecosystems.