<  Back to videos

LiveAI - Defending Against Adversarial AI - Spencer Thellmann - Palo Alto Networks

January 23, 2025

Coming in hot with a special episode of LiveAI to tide you over until season two officially launches! Your LiveAI host Tony Olzak couldn't wait for season two to find more emerging AI advantages, so he sat down with Spencer Thellmann to learn how he's defending adversarial AI with AI at ‪@paloaltonetworks‬. Their discussion includes the rise of prompt injection attacks, the process of developing a defense against something non-deterministic in nature, why it's essential to look at what users are asking your AI app, and much more!

After listening in, let us know if you want an invite to Tony and Spencer's GenAI-planned dinner outing. We'll get you on the list!

You might also enjoy

AI Readiness & Data Agility - Solving the Data Explosion with Trace3 & Cribl

Is an avalanche of telemetry data slowing down your AI initiatives? In this episode from Trace3's PressBox episode, experts from Cribl and Trace3 discuss how to gain data agility, reduce costs, and prepare for agentic AI environments. Trace3's Josh Lindstrom and Kevin Donahoe sit down with Michael Contreras from Cribl to cover everything trending in data and observability for 2026.

[00:48] Preparing Telemetry Data for AI

[02:14] Observability of the Data vs. Using the Data

[04:47] Case Study: De-risking a Global SIM Migration

[06:31] Supporting Agentic AI Environments

[08:27] The Trend of Data Ownership and Agility

[10:38] Reducing Overlapping Tool Costs & Waste

[12:48] Sneak Peek: Cribl Search 2.0 and Natural Language Queries

AI is a Data Security Problem: How to Secure the Autonomous Enterprise with Sentra & Trace3

Is your organization ready for AI, or is hidden 'shadow data' creating massive security risks? In this Trace PressBox discussion, Trace3 Advisory CISOs Gina Yacone and Jim Hunter dive deep with Sentra's Co-Founder Yoav Regev into why data security is the heart of the AI challenge. Check out these key moments:

[00:48] Why AI is a Data Security Problem

[02:14] The 3 Pillars: Discovery, Classification, and Context

[05:17] Complementing Microsoft 365 and Copilot

[08:06] AI as a "Risk Magnifier"

[11:13] Scaling Data Security for the Enterprise

[13:17] Requesting a Proof of Value (POV)

[15:21] Identifying Shadow Data and Reducing Cloud Costs

Future of App Security: Why the Trace3 + F5 Partnership is Critical for the Autonomous Enterprise

How do you maintain application performance across a fragmented multi-cloud footprint? In this Trace3 PressBox episode, we sit down with F5 to discuss application security, delivery, and the shift toward automated infrastructure.

0:00 - Introduction to the PressBox

1:30 - The Current State of Application Security

3:45 - Why F5 is Moving Beyond Load Balancing

6:20 - Solving Multi-Cloud Complexity

9:10 - Future Outlook: AI and App Delivery

11:00 - Closing Thoughts & Resources