The rapid deployment of autonomous AI agents is transforming enterprise operations, but it also creates new, unsecured attack surfaces within core infrastructure. AppGate, a leader in Zero Trust Network Access (ZTNA), has launched Agentic AI Core Protection, a new capability designed to secure these machine-driven workloads at scale. This innovation enables organizations to embrace AI-driven innovation while maintaining robust security and compliance by extending Zero Trust principles to AI agents and automated processes.
AppGate has launched Agentic AI Core Protection, a new capability within its AppGate ZTNA platform.
It is designed to secure AI workloads and agents deployed in servers, virtual machines (VMs), and Kubernetes clusters.
The solution addresses the security gap left by traditional ZTNA and VPNs, which focus on human users and leave non-human identities and machine-to-machine traffic unprotected.
Key features include a Linux Headless Client for servers/VMs, Kubernetes integration for pod-level isolation, and dynamic, context-aware access policies.
The technology cloaks infrastructure using Single Packet Authorization (SPA), making AI services invisible to unauthorized entities.
Agentic AI Core Protection is generally available as part of AppGate ZTNA as of December 18, 2025.
As enterprises accelerate AI adoption, autonomous agents are increasingly deployed directly into core environments like servers, VMs, and Kubernetes clusters, often to meet specific performance or compliance requirements. These deployments frequently expose APIs and web interfaces, creating new attack surfaces that traditional security models cannot address. Legacy VPNs and even conventional ZTNA solutions are primarily built to secure access for human users, leaving critical gaps for machine identities and automated traffic. This exposes organizations to risks such as lateral movement, unauthorized access, and compliance violations.
Agentic AI Core Protection directly tackles this challenge by applying Zero Trust security principles at the workload and machine level. “AI agents are transforming enterprise operations - but they also open new attack surfaces. Extending Zero Trust to the network layer is critical,” said Leo Taddeo, CEO at AppGate. “Agentic AI Core Protection locks down AI agents at the core, so organizations can innovate at scale without compromising security and compliance”.
The new capability integrates seamlessly into existing AppGate ZTNA, providing identity-based security and creating micro-perimeters around AI workloads in both on-premises and cloud environments. Its architecture is built for the unique demands of AI operations:
Linux Headless Client: Enforces ZTNA policies on servers and VMs without a graphical user interface, which is ideal for headless AI workloads.
Kubernetes-Native Integration: Offers sidecar and node-level enforcement to provide granular, pod-level isolation for dynamic containerized AI agents.
Dynamic, Context-Aware Policies: Access is granted based on a combination of identity, real-time device/workload posture, compliance state, and other risk signals, not just static network locations.
Infrastructure Cloaking: Leverages Single Packet Authorization (SPA) to make AI services, APIs, and dashboards completely invisible to network scans until the requesting entity is authenticated and authorized.
The primary business outcome of Agentic AI Core Protection is to allow organizations to deploy and scale AI rapidly without exposing their core infrastructure. By ensuring only authorized agents and users on compliant systems can access specific services, it reduces the attack surface and enforces least-privilege access. This approach not only improves security but also aids in regulatory compliance by providing clear audit trails for all machine-to-machine interactions.
The solution is positioned as essential for modern enterprises, where AI experimentation and deployment are often decentralized. As noted in an industry analysis, securing AI agents is not about discouraging innovation but about making robust security attainable so it supports the work instead of becoming a burden that gets ignored.
By extending its proven ZTNA framework to the unique challenges of autonomous AI, AppGate provides a path for organizations to secure their most valuable assets and data while fully leveraging the transformative power of artificial intelligence.
About AppGate
AppGate secures and protects an organization’s most valuable assets with its high performance Zero Trust Network Access (ZTNA) solution and Cyber Advisory Services. AppGate is the only direct-routed ZTNA solution built for peak performance, superior protection and seamless interoperability. AppGate safeguards enterprises and government agencies worldwide.
TechIntelPro empowers B2B technology decision-makers with the latest industry trends, deep insights, leadership perspectives and actionable intelligence to drive both organizational and professional milestones—and thrive on the tech pulse.