Q-Cyber Magazine

Enterprise · 18 June 2026 · 3 min read

Zero trust for the enterprise: Least privilege, requested each time

NIST SP 800-207 and CISA's Zero Trust Maturity Model 2.0 give enterprises a way to retire implicit trust on corporate networks, one resource at a time.

Abstract graphic of concentric signal rings in Knowlcom blue

Most enterprise networks still grant a great deal of trust once a user or device is "inside." Zero trust is the decision to stop doing that, and to make each access request earn a fresh, narrow authorisation.

The corporate network is no longer the argument

Enterprises adopted remote work, personal devices and cloud services faster than they retired the habit of treating the office network as a safe place. NIST's definition is a direct response to that mismatch. A resource should be protected because of what it is, not because of the subnet it occupies. A request from a staff laptop on a virtual private network should be examined with the same seriousness as a request from the public internet.

That does not make the network irrelevant. It changes the network's job from being a proof of trust to being a place where policy is enforced: segmentation, encryption, inspection, and the removal of unnecessary paths between systems. CISA's maturity model is useful here because it prevents a single-product reading of the idea. Identity, device posture, network path, application and data each have work to do, supported by visibility, automation and governance.

Exhibit 1

Core public references for an enterprise zero-trust programme

  1. August 2020

    NIST SP 800-207, Zero Trust Architecture, published as final

  2. September 2021

    CISA Zero Trust Maturity Model version 1.0 opened for public comment

  3. January 2022 onward

    CISA ZTMM 2.0 aligned to OMB memorandum M-22-09, with four maturity stages across five pillars

  4. May and October 2023

    Updated #StopRansomware Guide includes zero trust architecture among prevention practices

Source: NIST CSRC, SP 800-207 publication record; CISA Zero Trust Maturity Model page; CISA #StopRansomware Guide and 23 May 2023 announcement.

Where enterprises should start

A complete transformation is not the first deliverable. The first deliverable is the retirement of implicit trust in one or two places where it is still the default. Standing privileged accounts, flat administrative networks, shared file platforms, and service-to-service credentials embedded in code are typical starting points. Each should receive an owner, a policy, and a per-request check that includes user or workload identity and, where applicable, device or workload posture.

CISA's four-stage language — traditional, initial, advanced, optimal — is a honest reporting device for boards. Most enterprises will be uneven across pillars. Reporting that unevenness is more useful than declaring the organisation "zero trust" after a tooling purchase.

Implications

  1. Protect resources, not only the perimeter. Every critical application and data store needs an access policy that does not collapse to "on the LAN."
  2. Join identity to device or workload posture. A valid password on an unhealthy endpoint is an incomplete decision.
  3. Use segmentation to enforce policy, not to pretend the inside is safe.
  4. Govern exceptions. Emergency access that never expires will quietly restore implicit trust.

What this means for enterprise

Zero trust is a programme with a maturity language, not a purchase. Enterprises that can show, for their most important systems, who was allowed to do what, on which device, and for how long, are already practising the idea. The rest is extending that standard until implicit trust has nowhere left to sit.

Sources

  1. NIST SP 800-207, Zero Trust Architecture
  2. CISA Zero Trust Maturity Model
  3. CISA #StopRansomware Guide