AI is changing everything about how attacks are launched, how defenses are built, and how quickly the gap between the two can grow. But adopting agentic security operations without the right foundational context is not just ineffective, it is dangerous. Across three sessions, we expose how adversaries are weaponizing AI to outpace defenders, unpack why so many agentic SOC investments are quietly failing due to poor context, and, in partnership with CrowdStrike, lay out exactly what it takes to build agentic operations that actually work. This is the series for security leaders who are done with the hype and ready for the hard truths, and a clear path forward.
This series is designed for security operations leaders, SOC managers, security architects, and IT leaders who are evaluating, deploying, or looking to mature agentic AI capabilities within their security programs. Security analysts and practitioners seeking to understand how agentic workflows will impact their day-to-day operations will also find this series highly valuable.
The cybersecurity threat landscape is evolving at an unprecedented pace, and AI is at the center of that transformation. Adversaries are leveraging AI-assisted tooling and autonomous agentic attack frameworks to orchestrate sophisticated campaigns faster, at greater scale, and with far less human involvement than ever before. This session brings frontline security experts together to examine how these advancements are reshaping the balance between attackers and defenders, and what modern security teams must do to keep pace.
How AI is lowering the barrier to entry for sophisticated attacks, from AI-generated phishing and deepfake social engineering to automated vulnerability discovery and exploit development
As organizations rush to invest in agentic AI for their security operations centers, many are discovering a costly gap between expectation and reality. The culprit is often not the technology itself, but the data and context feeding it. Unlike a human analyst who might pause and flag that something doesn't add up, an AI agent won't tell you when it's missing critical context. It will act anyway, and it will act with confidence. When an agentic SOC operates with incomplete, siloed, or low-quality data, autonomous decision-making that looks authoritative but is built on an incomplete picture quietly erodes the trust, efficiency, and ROI that organizations were counting on. This session examines why so many agentic SOC investments are underperforming, and what organizations must address to realize the true promise of autonomous security operations.
Why AI agents don't know what they don't know, and why that blind confidence is one of the most dangerous characteristics of agentic systems operating in high-stakes security environments with incomplete data or context
How data quality issues, visibility gaps, and context fragmentation silently undermine agentic SOC performance, turning missed inputs into confident but incorrect autonomous actions that create more risk than they resolve
The promise of an agentic SOC is real, but so are the pitfalls. Having explored the evolving AI threat landscape and the critical role that data quality and context play in agentic performance, this session brings the series to a close with something actionable: a practical, practitioner-led blueprint for building an agentic SOC that actually works. In partnership with CrowdStrike, we will move beyond the theory and share the frameworks, lessons learned, and real-world guidance security teams need to deploy agentic operations with confidence, ensuring their AI has the visibility, context, and guardrails required to deliver on its promise.
The foundational requirements for an effective agentic SOC, including the data integrations, telemetry coverage, and context layers that must be in place before autonomous workflows can operate reliably
