AISPM Isn't Enough: How to Apply Zero Trust to AI Agents

About this episode

The episode discusses the importance of applying Zero Trust principles to AI security in the context of diverse AI tools and workflows.

We are officially entering the "Multi-AI Era." Much like the multi-cloud times, organizations are no longer just using a single AI tool like Microsoft Copilot, they are building custom, agentic workflows using diverse third-party models and MCP servers . In this episode, Ashish sits down with Shawn Hays from Varonis to discuss why the security market has over-pivoted on AISPM (AI Security Posture Management) . Shawn spoke about how having visibility and an inventory of your AI models is a great start, but it fails to secure the enterprise if you lack the guardrails to actually stop an agent from going off the rails and exfiltrating data . Shawn breaks down the components of a robust AI security platform (like Varonis Atlas) and explains why data security is inseparable from AI security. He spoke about why AI agents will blindly "read whatever is on the teleprompter," meaning your AI is only as secure as the data access and identity controls surrounding it . Tune in to learn how to apply Zero Trust across the entire AI chain from the prompter to the cloud infrastructure Guest Socials -⁠⁠ Shawn's Linkedin Podcast Twitter…

People in this episode

Host: Ashish

Guest: Shawn Hays

Topics covered

• Zero Trust
• AI Security
• Data Security
• AI Agents
• Security Management

Keywords

• Zero Trust
• AI Security
• Data Exfiltration
• AI Models
• Security Posture Management

Mentioned in this episode

Organizations: Varonis

Products: Microsoft Copilot, Varonis Atlas

More episodes of Cloud Security Podcast

• Native Cloud Firewalls Falling Short in a Multicloud World · June 11, 2026 · 37 min
• How AI Agents Will Negotiate Your Vendor Contracts · May 27, 2026 · 38 min
• How Claude Mythos Changes Vulnerability Management: From CVSS to Exploitability · May 5, 2026 · 45 min
• The Rise of Agentic Cloud Security: Code-to-Cloud Shrinks to 3 Days · April 21, 2026 · 27 min
• Why EDR Fails at AI Security & The Rise of Endpoint Behavior Modeling · April 14, 2026 · 31 min
• Solving Prompt Injection & Shadow AI for AI Malware · April 7, 2026 · 37 min