VERITRAILBook Governance Baseline

Enterprise AI governance, runtime guardrails and evidence

Govern every AI system, agent, and decision with evidence.

VERITRAIL Enterprise helps organizations inventory AI, assess risk, govern the lifecycle, protect LLM applications and agents at runtime, and prove responsible AI continuously.

Start the self-assessmentBook a Governance Baseline

VERITRAIL supports readiness and evidence. It does not certify compliance. Open the governance workspace.

An evidence trail — baseline, register, control and shield — leading to assured AI under the VERITRAIL shield, surrounded by inventory, maturity, policy and evidence-log cards.1 Baseline2 Register3 Control4 Shield5 AssuranceAI system inventoryHighMedLowMaturity index3.4 / 5Policy v1.3AllowDenyApproveEvidence log#a3f9 → #9c1e#9c1e → #42bd
  • 64governance controls
  • 5framework pillars
  • 6layers, one system
  • 24question self-assessment

Indicative mappings toNIST AI RMFEU AI ActISO/IEC 42001

Practitioner workspace · NIST AI RMF mapping

TILLER assessment & assurance studio

Open workspace
TTrust & AccountabilityNIST · GOVERNIInventory & OversightNIST · MAPLLifecycle & Use-Case ControlNIST · MAPLLawfulness, Ethics & ComplianceNIST · MEASUREEEngineering Reliability & GuardrailsNIST · MEASURERRuntime Operations & MonitoringNIST · MANAGE

Get started

Four ways to start

Why it matters

The problem: five questions most enterprises cannot answer with evidence

AI adoption has outpaced the governance designed to control it. Models arrive through vendor products, business teams build assistants on public APIs, and engineering teams give agents access to tools and data. When a board member, auditor, customer or regulator asks a direct question, the honest answer is often "we think so".

  1. What AI do we actually run? Which systems, models, agents and vendor AI features are in use, who owns each one, and what data does it touch?
  2. Which of them are high-risk, and why? Is there a consistent, approved basis for classifying decision impact, data sensitivity, autonomy and regulatory relevance?
  3. Who approved each system, and under what conditions? Can we show the risk assessment, the approver and the conditions that were attached before go-live?
  4. What stops an agent from doing something it should not? Are tool permissions, data handling and high-impact actions governed at runtime, or only described in a policy document?
  5. Could we produce evidence tomorrow? If asked, could we hand over an evidence pack for a material AI system — logs, approvals, test results, monitoring — without a fire drill?

VERITRAIL exists to turn each of those questions into a governed, evidenced answer.

The system

Six layers, one governance system

Six stacked layers forming one governance system.1 Baseline2 Register3 Control4 Shield5 Assurance6 Copilot

  1. Baseline

    Establishes where you are today, with evidence status for every finding

    VERITRAIL Assess and the maturity self-assessment

  2. System of record

    One owned, classified inventory of AI systems, models, agents and vendor AI

    VERITRAIL Register

  3. Lifecycle control

    Risk-proportionate intake, approval, change, monitoring and retirement gates

    VERITRAIL Control

  4. Runtime protection

    Policy-as-code guardrails for prompts, responses and agent tool calls

    VERITRAIL Shield

  5. Assurance

    Control testing, tamper-evident logs, incidents and audit-ready evidence packs

    VERITRAIL Assurance

  6. Copilot

    A governance agent that drafts assessments, gap registers and reports with citations

    VERITRAIL Agent

The framework

Five pillars, 64 controls

The VERITRAIL Framework organizes governance into five pillars and a library of 64 controls, each with an objective, an owner role, the evidence that demonstrates it, and indicative mappings to NIST AI RMF, ISO/IEC 42001 and the EU AI Act.

64 controls across 5 pillars; 17 are critical priority.AO12AO-01 AI governance policy (critical)AO-02 Executive accountability and AI governance forum (critical)AO-03 AI roles, responsibilities and RACI (high)AO-04 AI risk appetite and tolerance (high)AO-05 AI literacy and role-based training (high)AO-06 Escalation paths and decision rights (high)AO-07 Regulatory and contractual obligations register (high)AO-08 Responsible AI principles and objectives (medium)AO-09 Third-party AI accountability (medium)AO-10 Prohibited and restricted AI use policy (critical)AO-11 Stakeholder engagement and feedback (medium)AO-12 Governance resourcing (medium)EI12EI-01 Enterprise AI inventory (system of record) (critical)EI-02 AI risk classification and tiering (critical)EI-03 Accountable owner per AI system (critical)EI-04 Intended purpose and context of use (high)EI-05 Model, prompt and component registry (high)EI-06 Data source and lineage registry (high)EI-07 Agent and tool registry (critical)EI-08 Third-party and vendor AI inventory (high)EI-09 Shadow AI discovery (medium)EI-10 Dependency and integration mapping (medium)EI-11 Inventory attestation and recertification (medium)EI-12 Use-case intake register (high)GL12GL-01 Intake triage workflow (high)GL-02 Pre-deployment risk and impact assessment (critical)GL-03 Risk-tiered approval gates (critical)GL-04 Change management and reassessment triggers (high)GL-05 Pre-release validation and testing (high)GL-06 Technical documentation (high)GL-07 Deployment conditions and go-live readiness (high)GL-08 Post-deployment monitoring (high)GL-09 Retirement and decommissioning (medium)GL-10 Exceptions and risk acceptance (high)GL-11 Vendor AI due diligence and contract controls (high)GL-12 Periodic review and recertification (medium)IS16IS-01 AI threat modelling (high)IS-02 Data governance and quality (high)IS-03 Privacy assessment and data minimisation (critical)IS-04 Sensitive data redaction and masking (critical)IS-05 Bias and fairness testing (high)IS-06 Safety and harm evaluation (high)IS-07 Accuracy and robustness testing (high)IS-08 Transparency and AI disclosures (high)IS-09 Human oversight design (critical)IS-10 Secure model and prompt supply chain (high)IS-11 Least-privilege tool access (critical)IS-12 Runtime guardrails and policy enforcement (high)IS-13 Human approval for high-impact actions (critical)IS-14 Prompt-injection and data-leakage testing (critical)IS-15 Kill switch and rollback (critical)IS-16 Rate, budget and resource limits (medium)SE12SE-01 Control testing programme (high)SE-02 Evidence repository and retention (high)SE-03 Governance metrics and leadership reporting (medium)SE-04 AI incident management and reporting (high)SE-05 Audit-ready evidence packs (high)SE-06 Corrective action and continuous improvement (medium)SE-07 Management review (medium)SE-08 Immutable action logging (critical)SE-09 Post-market monitoring plan (high)SE-10 Independent review and internal audit (medium)SE-11 Stakeholder transparency reporting (medium)SE-12 Model performance and drift monitoring (high)
Critical High Medium priority

Browse every control in the VERITRAIL Control Library.

The method

How it works

  1. Baseline
  2. Register
  3. Control
  4. Shield
  5. Assurance
  1. Baseline. Start with the self-assessment, then validate it in a Governance Baseline. Every finding carries an evidence label — Verified, User-asserted, Inferred, Missing evidence or Critical gap — so leaders can see what is known and what is assumed.
  2. Register. Build the AI system of record: owners, intended purpose, data sources, models, agents, tools and vendors, each classified into a risk tier that a human confirms.
  3. Control. Route every new use case and material change through risk-tiered intake, assessment and approval, with exceptions that expire.
  4. Shield. Put versioned runtime policy in front of LLM applications and agents: redact sensitive data, deny unregistered tools, and require human approval for high-impact actions. Roll out in stages — Observe, Warn, Protect, Enforce — with a tested Contain mode for incidents.
  5. Assurance. Test controls, retain tamper-evident logs, manage AI incidents and produce evidence packs on request.

Our position

What VERITRAIL does — and does not — claim

VERITRAIL supports governance readiness, control implementation and evidence collection. Our framework mappings to NIST AI RMF, ISO/IEC 42001 and the EU AI Act are indicative interpretations designed to help you organize evidence. They are not an official crosswalk, a legal opinion or a conformity assessment.

Who it serves

Value by stakeholder

  • CISO

    What they need
    Visibility and control over LLM apps, agents and AI data flows
    What VERITRAIL provides
    Agent and tool registry, threat modelling, prompt-injection testing, runtime policy, kill switch and tamper-evident logs

  • CIO / CTO

    What they need
    Scale AI safely without slowing delivery
    What VERITRAIL provides
    Risk-proportionate gates, so low-risk use cases move quickly and high-risk ones get the right scrutiny

  • Chief Risk / Compliance Officer

    What they need
    A defensible risk position and readiness evidence
    What VERITRAIL provides
    Risk appetite linkage, tiering, exception register, indicative framework mappings and audit-ready evidence packs

  • Head of AI CoE

    What they need
    A repeatable intake-to-production path
    What VERITRAIL provides
    Standard intake, triage, assessment modules, approval workflow and reusable runbooks

  • Engineering lead

    What they need
    Clear, testable requirements instead of policy prose
    What VERITRAIL provides
    Policy-as-code with ALLOW / DENY / REQUIRE_HUMAN_APPROVAL rules, staged enforcement and change control

  • Consultant / advisor

    What they need
    A structured, evidence-first method for client work
    What VERITRAIL provides
    The control library, gap-analysis method, templates and the VERITRAIL Agent for drafting with citations

Next step

Start with an honest baseline

The self-assessment takes about 8 minutes and returns an indicative maturity level, domain scores, critical flags and a 30/60/90-day action plan. Results are self-reported and labelled User-asserted until evidence is reviewed.

Frequently asked questions

Does VERITRAIL make my organization compliant with the EU AI Act or ISO/IEC 42001?
No. VERITRAIL supports governance readiness, control implementation and evidence collection. It does not certify ISO/IEC 42001 conformity, determine EU AI Act compliance or replace legal advice. Framework mappings are indicative.
Where should we start?
Most organizations start with the free 8-minute maturity self-assessment, then validate the result with evidence in a 4–6 week VERITRAIL Governance Baseline.
Does VERITRAIL work with agents as well as models?
Yes. VERITRAIL registers agents and their tools, and VERITRAIL Shield evaluates prompts, responses and tool calls against versioned runtime policy, including human approval for high-impact actions.

Last reviewed: