Gaurav Duggal

Session Abstract:
As we stand in early 2026, agentic AI is no longer a futuristic concept—it’s actively transforming how enterprises operate, from hyper-automated SOCs that reduce analyst workload by up to 90% to multi-agent workflows executing complex tasks at machine speed. Yet, this extraordinary leap forward brings profound risks that innovation conversations too often overlook.

In my book Agentic AI at Risk: Cybersecurity Challenges and Laws in the Era of Autonomous Agents, I explore how these independent, goal-oriented systems introduce unprecedented vulnerabilities: agents as powerful “autonomous insiders” susceptible to goal hijacking, privilege escalation, and coordinated attacks that outpace human response; massively expanded attack surfaces where agents vastly outnumber people; and critical gaps in accountability, digital trust, and existing legal frameworks.

Drawing from real-world deployments at scale and emerging 2026 trends—like the rise of the “agentic SOC,” shadow autonomy, and AI-driven offense-defense imbalances—this talk cuts through the hype to offer a candid, practical perspective. We’ll examine the shift from prevention to resilience, the urgent need for enforceable governance, and strategies for developers, CISOs, and policymakers to build trustworthy agentic ecosystems.

Join me to understand not just the promise of autonomous intelligence, but how we must secure it before it secures (or compromises) us.

Target audience: Students, early-career professionals, founders, product managers, engineers, and business leaders.