Evidence for funders, insurers and customers

Someone has asked how your AI is governed. Here is what they will accept.

A lender, an insurer, a prime contractor or an enterprise buyer wants to know your AI is under control. A policy document and a questionnaire answer tell them what you intend. What moves the deal is evidence their own reviewer can check.

The AI Use Assessment →Book a conversation

Who asks, and what they accept

The question differs by requester. The test is the same.

Can the evidence be produced, and can someone independent check it?

Who asks	What they are asking	What they will accept
Lenders and project financiers	Whether the AI that generates the cash flow, or runs the asset, stays inside policy. Today the answer is the borrower's own compliance certificate.	Evidence they can monitor, attached to the action, that the borrower cannot quietly switch off.
Insurers and brokers	What the system was allowed to do, and what it actually did. Some policies now exclude AI until that can be answered.	An independent record, carrier-neutral, that any insurer can read and check.
Defence primes and global buyers	Supplier terms that flow their own obligations down to you, including what your AI may and may not do.	A technical schedule behind each clause: what is measured, how often, and against which limit.
Enterprise procurement	AI risk, model lineage and drift-governance questionnaires, often hundreds of questions long, before a contract is signed.	Answers their reviewer can check for themselves, rather than assurances they have to take on trust.

From a promise to a proof

Four conditions turn AI governance into something a third party can rely on.

A control below the operator

A mechanism the operator cannot quietly disable.

A record that travels

Evidence a third party can inspect, attached to the action itself.

A term drawn at the control

The covenant, policy condition or supplier clause references the mechanism, not a declaration.

Experience that accrues

A stream of measured outcomes, so terms can be priced on evidence over time.

The assessment, scoped to the requester

Start from the questions you have actually been asked.

The AI Use Assessment registers where AI acts in your product and operations, including inside your suppliers, and records what evidence each boundary can produce. Scoped to the requester's questions, it ends in a technical schedule their reviewer can check.

You provide

The questionnaire, covenant or supplier terms you have received
Your current AI inventory
The agreements with the suppliers in scope

No customer data, no system access.

You receive

A register of where AI acts, including inside suppliers
The evidence obtainable for each item the requester asked about
A finding per arrangement, and who owns each gap
A technical schedule the requester's reviewer can check

Three to four weeks, fixed fee, run through your counsel. Available now.

What moving first gives a supplier

Much of the value is defensive. Some of it wins work.

Sets the template

Requesters are still deciding what to ask. The first suppliers to answer with evidence shape the question.

Works at your pace

Wrong assumptions are corrected cheaply, before a tender or renewal deadline sets the pace.

Has the record first

What was asked, found and decided is recorded before a customer, insurer or regulator asks.

Wins the next tender

In defence and government tenders, and with global customers, evidence comes first and claims second.

Carrier-neutral by design

Your evidence stays yours.

Where controls produce records, they use an open, royalty-free format that any insurer, lender or customer can read and verify. No single vendor or underwriter can lock the evidence in.

The Universal Governance Protocol →