AI governance. Human accountability.

Govern AI systems.
Control access.

Connect AI ownership and risk decisions with technical controls. Evaluate supported MCP and HTTP actions, broker scoped access, and retain the evidence for review.

Request a demo See how it works

Your infrastructure. Your deployment.

Runtime action pathMCP / HTTP

Illustrative architecture

AI agentScoped token

Inline proxyPolicy evaluation

Technical policyIRON

Shadow mode

Downstream credentialsBrokered access

Decision recordHash-chained ledger

Credential brokerScoped access

Connected toolsSupported paths

Shadow mode observes policy decisions. Active enforcement requires configuration.

Governance decisions and runtime configuration stay distinct.

One governance program. Every kind of responsibility.

AI systemsHuman oversightTechnical controlsReviewed evidence

The governance workspace

A clear line from
responsibility to evidence.

Bring systems, people and decisions into the same view. Keep human controls as visible as technical ones.

Know what you govern.

Record AI systems, their owners and intended use. Keep scope and exact versions visible as the system changes.

Customer support assistantIllustrative workflow

Owner: System owner
Control: Human review
Review: Independent review

Put judgment in the right hands.

Use your risk method, decision authority and independent reviewers. Keep approvals tied to what was reviewed.

AssessReviewDecide

Evidence with context.

Connect attestations, tests and observations to the exact control. Keep collection and acceptance distinct.

Explore the evidence

Policy at the point of action.

Evaluate supported agent actions against IRON policies. Observe first, then apply the configured enforcement mode.

MCP / HTTPIRONDecision

Explore runtime controls

IRON

Explore the six IRON primitives

Evaluate supported agent actions against IRON policies. Observe first, then apply the configured enforcement mode.

The question it answers

Which agent acted, and what attribution is available?

What Rinox enforces

Scoped agent identity with device and human attribution where the source supplies it.

From intention to oversight

Every decision
has a next step.

Start with one AI use case. Follow its ownership, controls and evidence through a reviewable workflow.

Talk through your use case

01
Define the system
Purpose, owners and versioned scope
02
Make accountable decisions
Requirements, risks and independent review
03
Describe the control
Implementation, responsibility and evidence method
04
Review the evidence
Source, scope, result and validity

Illustrative governance journeyRINOX / 01

From intended use to reviewed evidence

Start with the system.

Purpose, owners and versioned scope

Purpose: Customer support assistant
Owner: System owner
Version: 1.0

Define the system before making a governance decision.

Evidence review

Illustrative workflow

Owner attestationControl owner

Control testExact control and system version

Runtime observationSupported action paths

A file receipt is not evidence acceptance.

A record you can examine

Keep the basis
for every decision.

Human review and runtime records answer different questions. Preserve their source, scope and review context without rewriting past approvals.

Versioned systems and control evidence
Named owners and independent review
Hash-chained runtime decisions

A shadow verdict is not a prevented action; an allow decision is not proof of execution success.

Supported runtime controls

Set the boundaries.
See the decisions.

Technical controls apply to connected, supported action paths. Governance approval does not automatically publish a runtime policy.

Shadow first, enforce when ready

Shadow mode records would-be decisions. Active enforcement and failure handling require explicit configuration.

IdentifyScoped action identity

Actions on supported paths carry an agent identity and, where supplied by the integration, device and human attribution. Credential brokering lets an agent use a Rinox-scoped token while Rinox holds the downstream key.

ConstrainDeny by default, down to the argument

An agent may use only the tools its policy names. An allowed tool can still be refused on its arguments: a path outside the workspace, a host off the list, a destructive statement matched by pattern.

LimitSession caps stop runaway use

Per-tool limits on calls and bytes per session. A looping agent or a bulk pull stops at the cap even when every individual call was allowed.

SequenceDangerous chains are denied as chains

Policies can deny an observed sequence, such as reading a sensitive file followed by sending data externally, optionally within a time window. Sequence checks apply to the actions Rinox receives.

ContextAnomalous behavior is escalated

An action that is unusual for this agent (a new tool, a new transition, drift from the declared task) is escalated to warn, hold for a human, or block. Escalation only: a deny never becomes an allow.

AttestRecorded decisions and approvals

Allow, warn, step-up, or block, each verdict is appended to a hash-chained ledger as it is made, with the policy that produced it. Held actions wait for a named approver and record who decided.

Framework context

Your obligations.
Your confirmed scope.

Work with named source editions and reviewed requirements. A reference mapping is a starting point, not proof of compliance.

ISO/IEC 42001

Reference mappings for AI management system controls. Organizational responsibilities and human review remain part of the governance work.

NIST AI RMF

References for organizing AI risks, controls, monitoring and review evidence.

EU AI Act

References for risk management, logging and oversight. Applicability and control sufficiency require review for the specific system and role.

Saudi requirements

English and Arabic control authoring for Saudi sources, including PDPL, SDAIA and NCA references. Framework content and source versions require qualified review.

Qualified review of the selected requirements and evidence is part of each engagement. Saudi framework content remains unverified until that review is recorded. Rinox does not certify an organization or guarantee regulator or auditor acceptance.

Built for your operating environment

Keep your
infrastructure choices.

Run on customer-controlled infrastructure, including on-premises and air-gapped environments. Agree the supported profile and acceptance criteria for your deployment.

Air-gapped and on-premises

The decision point, policy store, ledger and secret store can run locally without cloud dependencies in the data path.

Single-tenant

Run the same components on infrastructure you control, with containers and deployment tooling.

Hosted, multi-tenant

Tenant-scoped records and database row-level security separate organizational data.

Customer-controlled infrastructure

Illustrative deployment

Decision point

Policy store

Ledger

Secret store

Local operation. Explicit boundaries.

Start with one AI use case

Move forward
with a clear basis.

Walk through ownership, risk decisions, supported controls and deployment fit with the Rinox team.

Request a demoNo customer datasets or sensitive documents needed.

01Your AI use case.

02Ownership
03Controls
04Evidence

Decisions you can trace.

Keep the basis
for every decision.

Human review and runtime records answer different questions. Preserve their source, scope and review context without rewriting past approvals.

Versioned systems and control evidence

Named owners and independent review

Hash-chained runtime decisions

A shadow verdict is not a prevented action; an allow decision is not proof of execution success.

Keep your
infrastructure choices.

Run on customer-controlled infrastructure, including on-premises and air-gapped environments. Agree the supported profile and acceptance criteria for your deployment.

Air-gapped and on-premises

The decision point, policy store, ledger and secret store can run locally without cloud dependencies in the data path.

Single-tenant

Run the same components on infrastructure you control, with containers and deployment tooling.

Hosted, multi-tenant

Tenant-scoped records and database row-level security separate organizational data.

Govern AI systems.Control access.

A clear line fromresponsibility to evidence.

Know what you govern.

Put judgment in the right hands.

Evidence with context.

Policy at the point of action.

Explore the six IRON primitives

Which agent acted, and what attribution is available?

Every decisionhas a next step.

Define the system

Make accountable decisions

Describe the control

Review the evidence

Start with the system.

Keep the basisfor every decision.

Set the boundaries.See the decisions.

Your obligations.Your confirmed scope.

Keep yourinfrastructure choices.

Move forwardwith a clear basis.

Govern AI systems.Control access.

A clear line fromresponsibility to evidence.

Know what you govern.

Put judgment in the right hands.

Evidence with context.

Policy at the point of action.

Explore the six IRON primitives

Which agent acted, and what attribution is available?

Every decisionhas a next step.

Define the system

Make accountable decisions

Describe the control

Review the evidence

Start with the system.

Keep the basisfor every decision.

Set the boundaries.See the decisions.

Your obligations.Your confirmed scope.

Keep yourinfrastructure choices.

Move forwardwith a clear basis.

Govern AI systems.
Control access.

A clear line from
responsibility to evidence.

Every decision
has a next step.

Keep the basis
for every decision.

Set the boundaries.
See the decisions.

Your obligations.
Your confirmed scope.

Keep your
infrastructure choices.

Move forward
with a clear basis.

Govern AI systems.
Control access.

A clear line from
responsibility to evidence.

Every decision
has a next step.

Keep the basis
for every decision.

Set the boundaries.
See the decisions.

Your obligations.
Your confirmed scope.

Keep your
infrastructure choices.

Move forward
with a clear basis.