When AI Agents Act Autonomously: Who Decides What They Are Allowed to Do? Autonomous Agents in a Zero-Trust Network: The Okta and netgo AI Test Case
As AI agents gain autonomy, a critical security question arises: Who defines what they are permitted to do? Based on a real-world test case by Okta and netgo, this article explores how Zero-Trust architectures can be leveraged to securely govern autonomous AI agents, dynamically verify identities, and strictly limit access permissions.
Topic
When & Where
Details
Format:
Technology lecture
Language:
German
Session description
Speakers:
Ismail Aksoy, Okta
Stephan Gonschorek, netgo
Stephan Gonschorek, netgo The deployment of autonomous AI agents is transforming corporate processes, yet it introduces novel security and governance challenges. When systems independently make decisions and execute actions across digital infrastructures, traditional role-based access control is no longer sufficient. This raises a fundamental question: Who controls the authority of digital entities, and how can misuse be effectively prevented?
Drawing on a joint practical test case by Okta and netgo, this paper analyzes how the core principles of a Zero-Trust network ("Never Trust, Always Verify") can be extended to autonomous AI agents. It demonstrates how agents can be managed as distinct identities, pro ...
