Neo-Security: One Security Model for Humans and AI Agents

Teams are shipping MCP servers and AI agents faster than they’re securing them.

The result is a growing set of LLM-powered components with broad access to enterprise data and no clear identity, privileges or audit trail.

In this session, we walk through a deployment model - Neo-Security architecture - that lets you secure APIs and new types of backend workloads with one approach for both humans and AI. You’ll see how to:

• Create AI zones that keep sensitive data private
• Associate every agent to a secure identity and the attributes to match
• Restrict what agents can do and trace who authorized each call across complex API flows
• Apply the interoperable standards that make agent collaboration safe.