Join our Newsletter — 33% off our NHI Course
Home› Glossary› Identity Beyond IAM› Google Zanzibar
Identity Beyond IAM

Google Zanzibar

← Back to Glossary
By NHI Mgmt Group Updated October 31, 2026 Domain: Identity Beyond IAM

Google Zanzibar is a distributed authorization model designed for large-scale, fine-grained access control. It represents permissions as relationships between users, groups, and resources, then evaluates access through a central service. The approach is widely referenced for systems that need expressive authorization at scale.

What Google Zanzibar Is

Google Zanzibar is best understood as a distributed authorization model for large systems, not a standalone product feature. It centralises access decisions while modelling who can do what through explicit relationships between users, groups, and resources. That makes it suitable for environments where permission logic must stay expressive, consistent, and scalable.

Its core value is that authorization becomes a first-class service rather than scattered logic embedded in every application. That separation helps organisations keep policy logic coherent as the number of apps, tenants, and resource types grows.

How Zanzibar Models Access

Zanzibar-style systems represent access as relationship tuples, such as a user being a member of a group or a group having access to a document. The system then evaluates whether a request should succeed by traversing those relationships through a central authorization engine. This approach is often described as relationship-based access control, or ReBAC.

The model is more expressive than simple role assignment because it can capture nested sharing, ownership chains, delegated access, and resource-specific exceptions. It is also more operationally demanding, because the correctness of the permission graph becomes part of the security boundary.

Why It Matters for Large-Scale Authorization

Zanzibar matters when organisations outgrow coarse-grained access control and need fine-grained decisions at high request volume. It is designed for cases where per-resource permissions, inherited sharing, and dynamic group membership must be checked reliably without hard-coding custom authorization rules in each service.

That centralisation improves consistency, but it also means the authorization service and its data model become business-critical dependencies. If the relationship graph is incomplete, stale, or inconsistently replicated, users may be denied access they should have, or granted access they should not.

For background on the broader design choices behind relationship-based access control, Authorisation Models Guide compares RBAC, ABAC, ReBAC and related approaches.

Operational Concerns and Common Failure Modes

In practice, the hardest problems are usually not the access checks themselves, but the lifecycle around them: tuple creation, change propagation, caching, consistency, and rollout safety. A Zanzibar-style system can be secure in principle and still fail if replicas lag behind, if revocation is slow, or if application teams mis-model the underlying relationships.

Because permissions are computed from a graph, seemingly small data errors can have wide blast radius. One incorrect group edge or missing revocation can expose large resource sets, especially when the system is used as the shared authorization layer for many applications.

The implementation details of tuple handling, consistency, and safe migration are discussed in Fine-Grained Authorisation Guide, which covers schemas, rollout concerns, and the new enemy problem.

Risk and Threat Considerations

Google Zanzibar concentrates authorization power into a small number of services and data paths, which makes integrity and consistency failures especially consequential. If an attacker can tamper with relationship data, exploit stale replicas, or abuse an application that trusts the authorization layer too broadly, the result can be unauthorized access at scale.

Failure mechanism: stale or inconsistent relationship state, overly permissive tuple creation, or broken object-level authorization in a dependent application can cause access decisions to diverge from intended policy.

Impact: the system may expose sensitive resources, preserve access after revocation, or create broad privilege errors that are difficult to detect because the authorization engine still appears to be functioning normally.

Standards & Framework Alignment

This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.

OWASP API Security Top 10 addresses the attack and risk surface, while NIST SP 800-53 Rev 5 sets the governance and control requirements practitioners need to meet.

FrameworkControl / ReferenceRelevance
OWASP API Security Top 10API1 — Broken Object Level AuthorizationZanzibar-style permissions directly aim to prevent object-level access failures.
Recommendation — Model object access through centralized authorization decisions to prevent broken object-level authorization.
NIST SP 800-53 Rev 5AC-3 — Access EnforcementZanzibar is an access-enforcement mechanism for fine-grained authorization decisions.
AC-6 — Least PrivilegeRelationship-based authorization supports minimal access by expressing narrowly scoped permissions.
AU-12 — Audit Record GenerationLarge-scale authorization services need traceable decision logging for review and investigation.
Recommendation — Enforce access centrally so every request is evaluated against the current authorization policy. Constrain permissions to the minimum relationships needed for each resource and action. Log authorization decisions and tuple changes so access paths can be reconstructed during review.

Practitioner Guidance

Why practitioners should care: Zanzibar is powerful precisely because it turns authorization into infrastructure, so governance must cover schema design, relationship ownership, and rollout safety. Treat the permission graph as security-critical data, not as implementation detail.

What to watch for: focus on revocation latency, cache staleness, unexplained access discrepancies, and application teams bypassing the central model for edge cases. Those are the conditions that usually turn a sound authorization design into a real exposure.

Practitioner takeaway: the strongest Zanzibar deployments combine expressive policy modelling with strict change control and continuous verification of relationship data.

Free weekly newsletter

Subscribe to the NHI & AI Identity Journal

The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.

Bonus 33% off our NHI Course when you subscribe.

NHIMG Editorial Note
Reviewed and updated by the NHIMG editorial team on October 31, 2026.
NHI Mgmt Group — the #1 independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org