The Defender's Advantage Podcast Agentic AI in Cybersecurity
9 snips
Feb 5, 2025 Steph Hay, Senior Director at Google Cloud Security, discusses the transformative role of agentic AI in cybersecurity. She explains how generative AI is already reducing threats and analyst toil while scaling workforce talent. Steph highlights the shift from summarization to automated agents that streamline security workflows. These semi-autonomous agents enable rapid threat assessments and improve response times by handling repetitive tasks. She emphasizes the importance of safety controls and collaboration with customers to refine AI capabilities.
AI Snips
Chapters
Transcript
Episode notes
Joining Security After A Breach
- Steph joined Google Cloud Security from financial services after experiencing a data breach in her former company.
- That breach motivated her focus on user experience and tooling to help defenders reduce toil and close visibility gaps.
Three T's: Threats, Toil, Talent
- Generative AI has been used to reduce threats, cut toil, and amplify talent in security workflows.
- The team shifted from summarization to automating multi-step workflows with agentive capabilities.
Agents Automate Goal-Oriented Workflows
- An agent models a human's multi-step workflow, identifies repeatable parts, and automates them toward a goal.
- In security, agents should be semi-autonomous, automating low-risk tasks and surfacing items needing human approval.