Home / Agent Passport™
The Agent Passport is the capability-based access model at the centre of the control system. Every person who uses AI in your organisation is passported to the level appropriate to their role and training. Access expands with demonstrated capability. Structured, not open by default.
01 / The tiers
Each passport level maps to defined permissions: what AI tools can be accessed, what can be built, and what requires governance approval. Passport levels are assessed before any permission change or use case approval is made.
Understands AI outputs and uses approved tools within defined parameters. No build capability.
Configures and applies AI within structured workflows. Can adapt approved tools to business context.
Designs and deploys AI agents and integrations. Subject to full governance, security review and sign-off protocols.
02 / Appointed roles
Passport levels recognise capability. Accountability is different: it is appointed. Two governance roles sit alongside the three levels and hold the decision rights that keep the system defensible.
Accountable for a use case or agent in production: its value, its operation, and its lifecycle. A named owner stands behind every live deployment.
APPOINTEDApproves risk, policy and compliance decisions at control gates. Sits on the governance forum. The named answer to the question regulators now ask: who approved this, on what evidence?
APPOINTED03 / In the lifecycle
The Agent Passport governs who can do what at every stage of the five-stage lifecycle. It is not a certification, a credential or a qualification. It is how access is decided inside the control system, applied at every governance gate through your existing tooling and approval workflow.
Passport framework introduced. Governance lead and capability lead identified.
INTROPassport levels defined and assigned to the pilot cohort. L1 Consumer baseline set for all staff.
ASSIGNPassport criteria confirmed for use case pilot participants. L2 Practitioner pathway opened.
OPENPassport compliance monitored. Rogue access flagged and escalated. L3 Builder access governed.
ASSESSPassport progression assessed via six-month capability reassessment. New cohorts onboarded.
PROGRESS"AI does not scale through access. It scales through control."
The passport principle04 / Where it starts
The AI Control Diagnostic includes a capability baseline for your workforce using the Agent Passport framework. You will know who can use AI, at what level, and to what effect, before anything scales.
Why capability-based access matters
Access is mapped to demonstrated capability, not seniority or enthusiasm.
Every permission change is assessed against passport level before approval.
Accountability is personal: every AI user holds a level with defined limits.
A repeatable six-month reassessment cycle tracks maturity progression.