A Practitioner’s Guide to Secure Against Threats from Mythos-like AI Models

Frontier AI models like Mythos have changed the threat landscape, autonomously discovering vulnerabilities, creating advanced attack paths by linking CVEs, and orchestrating attacks at machine speed. The question isn't if adversaries will weaponize these capabilities, but how prepared you'll be when they do.

Drawing on work as part of Anthropic’s Project Glasswing and OpenAI Daybreak, senior security leaders at Zscaler will provide you with a practical playbook you can act on immediately. Whether you are just getting started with AI threat readiness or refining controls you already have, this session will dive into how to leverage Zscaler to protect your organization.

You'll learn how to:

• Reduce attack surface with Zero Trust segmentation
• Execute a security policy assessment, tailored for Mythos
• Discover your AI assets and run AI red teaming
• Deploy deception to catch AI-driven attacks
• Leverage vulnerability management and exposure management