Thursday, October 16, 2025 / 11:45 AM -
12:30 PM at Juniper 1, West Convention Center, Level 3
Zero trust for AI: Securing Model Context Protocol communications
This session delivers a practical framework for implementing zero trust principles in AI model interactions using Model Context Protocol (MCP). Discover how Cloudflare Zero Trust secures the critical connections between AI applications, models, and data sources. See real-world implementations for authenticating model access to sensitive data, enforcing least-privilege principles for AI agents, and creating secure boundaries between AI components. The session includes specific configurations for implementing identity-based policies, establishing secure transport for MCP communications, and creating comprehensive audit trails of all AI interactions across your environment.
Key Takeaways
- Learn how to implement zero trust principles for MCP to secure interactions between AI models and sensitive data while maintaining performance.
- Discover techniques for authenticating, authorizing, and auditing AI model access to tools and information sources across multicloud environments.
- Gain strategies for implementing granular, contextual access controls for AI applications using Cloudflare Access and Gateway.
Speakers
