Zero-trust security startup Zero Networks has launched an AI Segmentation feature to control AI agents within companies based on identity and block AI-enabled lateral movement across networks, SiliconANGLE reported.

As AI agents surge in corporate environments, 3 security gaps are emerging: a lack of controls over agents accessing sensitive systems, lateral movement accelerated by AI-driven attacks, and compliance operations that cannot keep up with the pace of change.

AI Segmentation offers 3 features to address this. AI Lateral Movement Control blocks connection paths that attackers or autonomous agents could use to reach critical resources, applying the principle of least privilege based on identity and network.

AI Agent Control identifies which agents are running in an environment, what they access and how they communicate, and applies strict boundaries to every interaction. It also detects shadow AI tools used without a company’s knowledge.

An AI-based compliance and risk engine lets security teams query billions of network activities in natural language and automatically map them to frameworks such as the EU NIS2 directive and CIS Benchmarks, assigning dynamic risk scores.

Benny Lakunishok (베니 라쿠니숍), co-founder and CEO of Zero Networks, said, "Zero Networks enables companies to fully control AI." He said it "highlights only what truly matters with real-time deterministic control over AI agents, AI-based visibility, and an integrated compliance and risk engine."