The Job Security Cybersecurity Podcast

Episode 3: Building an AI-powered security practice

Expel MDR Season 1 Episode 3

Host Dave Johnson and co-host Tyler Zito sit down with Peter Holcomb, founder and CEO of Optimo IT and self-described "AI Samurai," to explore how AI is reshaping cybersecurity—from automating compliance workflows to defending against emerging threats. Peter shares practical insights on shadow AI risks, AI observability, and how fractional CISOs are becoming essential for AI-native companies navigating security and governance challenges.

Key topics & timestamps

Peter's background and Optimo IT (2:31 - 4:26)

  • Founder/CEO of AI security consulting specializing in fractional CISO services
  • Focus: SOC 2 Type II, ISO 42001/27001, GDPR, HIPAA
  • Former CISO at DataVolo (acquired by Snowflake) and EMED Digital Healthcare

Overlooked AI security challenges (4:26 - 7:35)

  • Shadow AI becoming the new "shadow IT"—unsanctioned tools introducing risk
  • AI observability must track: alert severity, user queries, token usage, cost, data lineage
  • Automated evidence tracking with platforms like Vanta, Drata, Risk 360

Applying existing security principles to AI (7:35 - 9:02)

  • Reapplying standard security practices to different use cases
  • Continual education on appropriate tool usage and data stewardship
  • Shared responsibility between security teams and business

The fractional CISO model (9:02 - 14:24)

  • AI-native companies need security expertise but want to focus on product
  • Business owns the risk—CISO advises on treatment options
  • Third-party perspective often carries more weight than internal recommendations

Building an AI-powered business (16:17 - 19:32)

  • Email agent automates responses, saves drafts for review
  • Lead generation agents personalize outreach sequences
  • ~10 agents handling administrative tasks to focus on strategic work
  • Building evidence collection agents for audit workflows

AI security use cases (19:32 - 24:21)

  • Red team/blue team testing via TestSavant.ai
  • Microsoft Copilot integration risks
  • Recommended tools: Petra Security, Cloud Capsule for pre-Copilot assessments

AI's future in security operations (24:43 - 28:27)

  • Near-term: Autonomous defense agents detecting/remediating faster than humans
  • Still need human-in-the-loop for verification
  • Zentra.ai: Building agents for level 1-2 IT operations
  • Example: 24-hour ticket resolved in 30 seconds with agent automation

Career advice (29:41 - 32:22)

  • Get educated on AI—tinker with it, understand pitfalls
  • AI governance is the "new GRC"
  • Get hands-on: Build labs, use AWS free tier, experiment with tools
  • Identify repetitive tasks and automate with agents

Key quotes

"Shadow AI is becoming a huge thing right now... individuals want to be more productive, but they might install these vibe coded tools and now they're introducing more risk into the environment." - Peter Holcomb

"There are only four things you can do with risk. You can accept the risk, mitigate the risk, transfer the risk, or ignore the risk." - Peter Holcomb

"Back in the day, GRC was not looked at as a sexy thing, but now, with the ubiquity of AI, AI governance is top of mind for everybody." - Peter Holcomb

Production Credits

  • Co-hosts: Dave Johnson and Tyler Zito
  • Producer: Ben Baker
  • Sponsor: Expel MDR

Connect