Zero Networks launched AI Segmentation on April 21, adding three capabilities to its zero-trust platform that explicitly target the security gaps created by autonomous AI agents in enterprise environments, according to SiliconANGLE.
The release addresses three specific problems: agents accessing systems beyond what they need, AI-accelerated lateral movement across corporate networks, and compliance teams unable to keep pace with agent proliferation.
Three Capabilities
AI Lateral Movement Control enforces identity and network-based least privilege for agent connectivity. It cuts off the paths that attackers or autonomous agents can use to reach critical resources, applying zero-trust segmentation at the network layer specifically for non-human identities.
AI Agent Control provides visibility into which agents are running inside an environment, what they are accessing, and how they communicate. This includes detection of shadow AI tools operating without IT awareness. The system enforces strict boundaries on every agent interaction, as described by SiliconANGLE.
AI-Powered Compliance and Risk Engine lets security teams query live network activity in natural language across billions of connections. It maps activity against the EU’s NIS2 Directive and CIS Benchmarks, assigning dynamic risk scores. This addresses the compliance velocity problem: regulatory frameworks don’t update as fast as agent deployments change network behavior.
“Zero Networks puts enterprises in control of AI, full stop,” CEO and co-founder Benny Lakunishok told SiliconANGLE. “Real-time, deterministic control over AI agents, combined with AI-driven visibility and an integrated Compliance and Risk Engine that continuously scores risk, maps activity to frameworks like NIS2 and CIS, and flags what actually matters.”
Funding and Availability
All three capabilities are available now in the Zero Networks platform. The company has raised approximately $100 million in total funding, including $20 million in December 2023 and $55 million in June 2025, according to SiliconANGLE. Investors include Highland Europe, F2 Venture Capital, PICO Venture Partners, Venrock Associates, and U.S. Venture Partners.
The Agent Security Stack Takes Shape
Zero Networks joins a growing list of security vendors building infrastructure specifically for agent governance. The consensus emerging from RSAC 2026 was that “AI agents must be governed like powerful insiders,” with identity-based segmentation, behavioral monitoring, and continuous compliance verification. Capsule Security exited stealth with $7 million for agent runtime security earlier this month. Aikido Security launched endpoint protection for developer-facing AI tools. Microsoft’s Agent 365 includes Purview-based observability reaching general availability in May.
The common thread: as enterprises deploy more autonomous agents, the security perimeter shifts from protecting human users to governing non-human identities that accumulate permissions, move laterally, and operate at machine speed.