Portnox Adds Network Kill Switch for Risky AI Agents

Portnox announced in a press release expanded access control capabilities for AI agents and other machine identities across enterprise networks, applications, and infrastructure. The update adds an integration with Microsoft Defender alongside existing integrations with CrowdStrike and SentinelOne.

Portnox says the integrations let its platform use risk signals from those security tools to apply access policies at the network layer. When a tool detects elevated risk, suspicious activity, or a device posture change, Portnox can block, quarantine, or revoke access for an identity, including an AI agent.

The company describes the feature as an access kill switch for identities that continue connecting after their risk status changes. The workflow detects risk through Microsoft Defender, CrowdStrike, or SentinelOne, evaluates it against Portnox policies, and enforces the required access change.

Portnox said the capabilities also give security and IT teams records showing which identity connected, when and where it connected, what it could access, and which policy governed the access decision.