Skip to content
Infilux AppSec Logo
Glossary · Zero Trust

What is Zero Trust?

Zero Trust is a security model that treats every user, device, and service as untrusted by default — regardless of whether they sit inside or outside the corporate network. Every access request is independently authenticated, authorised, and policy-evaluated based on identity, device posture, location, and context. The phrase 'never trust, always verify' captures the model. It replaces the legacy castle-and-moat perimeter with continuous verification at every resource.

Traditional perimeter security assumed: inside the firewall is trusted, outside is not. Once an attacker breached the perimeter (or an employee did anything on a VPN), they had broad lateral access. Cloud workloads, SaaS apps, remote work, and supply-chain attacks have made this assumption fatal. Zero Trust replaces it.

Implementation is layered. Identity becomes the new perimeter — strong authentication (MFA, phishing-resistant) plus device posture (managed? encrypted? patched?) plus risk signals (impossible travel, anomalous behaviour). Access is granted just-in-time, just-enough, to specific resources for specific durations — not 'VPN onto the network and reach everything.' Microsegmentation prevents lateral movement; identity-aware proxies replace VPNs; service-to-service authentication uses mTLS or signed JWTs rather than IP allowlists.

Zero Trust is not a product. It's an architecture pattern with multiple implementation paths — Google's BeyondCorp, Microsoft's Zero Trust framework, NIST SP 800-207. Most enterprises adopt incrementally: identity-aware proxies for SaaS first, then conditional access for managed devices, then microsegmentation for the highest-value workloads, then service mesh for east-west traffic.

Key points

  • Principle: 'never trust, always verify' — no implicit trust from network location.
  • Components: strong identity (MFA), device posture, just-in-time access, microsegmentation, continuous evaluation.
  • Replaces VPN-based perimeter with identity-aware proxies.
  • Reference architecture: NIST SP 800-207.
  • Adopted incrementally — most teams take 18–36 months for a comprehensive rollout.

Frequently asked

Is Zero Trust the same as ZTNA?+
ZTNA (Zero Trust Network Access) is a specific category of products that implement Zero Trust principles for remote access — replacing the VPN. Zero Trust is the broader architecture; ZTNA is one component. Other Zero Trust components include identity providers (Okta, Entra ID), CASB / SSE for SaaS, microsegmentation tools, and service-mesh-based service-to-service authentication.
Can we adopt Zero Trust without ripping out the VPN?+
Yes, and most enterprises do. The typical migration: deploy ZTNA alongside the VPN for new applications; gradually move existing applications behind the ZTNA proxy; retire the VPN for general access after 12–24 months; keep a small legacy VPN for niche cases (network appliance management, specific legacy apps).
What about service-to-service traffic in microservices?+
Same Zero Trust principles apply. Service-to-service authentication should use mutual TLS (mTLS) or signed-JWT bearer tokens, not IP allowlists. Service mesh tools (Istio, Linkerd, Consul) make this practical at scale by automating certificate issuance and rotation.