Let AI help you lend. Keep every approve, price, and decline authorised, explainable, and replay-provable.

The problem

Automated credit decisions are regulated, consequential, and hard to defend after the fact.

• An adverse action needs specific reasons. ECOA / Regulation B requires a creditor that declines or worsens terms to state the specific principal reasons for it.
• A credit-report decline triggers a notice. The FCRA requires a user who takes adverse action based on a consumer report to tell the borrower and name the reporting agency.
• The model must be governed. Fed SR 11-7 expects a model driving a decision to be validated and used only within its approved scope.
• UK lending must deliver good outcomes. The FCA Consumer Duty and CONC require good outcomes, consumer understanding, a creditworthiness assessment, and forbearance in arrears.
• The usual evidence is weak. Screenshots and reconstructed reason codes are contestable. Regulators and borrowers want a signed, verifiable record produced at the moment of the decision.

Why buy this

A signed Evidence Pack™ on every decision — verifiable from public keys alone.

• Named authority before the decision. No approve, decline, price, or arrears action runs until a purpose-scoped named-authority decision resolves. Unauthorised actions are refused.
• A Decision Map™ on every adverse action. An adverse credit action without a Decision Map™ — the specific principal reasons, explainable — is refused at the boundary. The reasons are sealed into a signed, replayable artefact.
• Model in scope, enforced. A model-driven decision proceeds only on a model the register shows validated and used within its approved scope.
• Every decline is contestable. An adverse action carries an explanation and a route for the borrower to contest it through the KYE™ rights-disputes rail.
• Honest scope. KYE™ governs the lending agent's authority-to-act and the adverse-action audit trail — not the credit model's statistical correctness or your fair-lending mathematics. That stays your own quantitative work.

How it works

Governance at the moment of the decision — across the whole credit lifecycle.

Every AI-assisted credit action flows through the same gate. The pipeline binds to canonical KYE Protocol™ envelopes — no one-off shapes per lender.

1. 1 — Action proposed. An AI agent proposes an underwriting, pricing, adverse-action, or arrears action over a consumer-credit application or account.
2. 2 — Authority check. The gate verifies a purpose-scoped named-authority decision and — for a report-driven action — an FCRA permissible purpose. No authority, no action.
3. 3 — Model in scope. A model-driven decision is checked against the model register for current validation and approved scope (Fed SR 11-7).
4. 4 — Decision Map™ for adverse actions. An adverse action must carry the specific principal reasons, explainable, before it is communicated.
5. 5 — Evidence Pack™ sealed. The decision emits a signed Evidence Pack™ binding the authority, the model reference, the Decision Map™, and the outcome.
6. 6 — Borrower contestability. An adverse action carries an explanation and a bound route to contest it, recorded on the envelope.
7. 7 — Quarterly attestation. The accountable owner co-signs a quarterly attestation covering the decision register, open contests, and model-validation currency.

Framework binding

Bound to the consumer-lending regulatory spine.

The Pack binds the canonical KYE™ artefact set to the regulations that govern consumer-credit AI. Every claim resolves to a control row on the bound framework.

Full coverage means every requirement in the control area is bound to a canonical KYE™ rule, dictionary term, or Evidence Pack™ contract. Out-of-scope areas — the disparate-impact statistics, the affordability calculation, the model's fairness validation — stay the lender's own quantitative work; coverage is never inflated to 100%.

Get started

Run a pilot on synthetic data — then deploy with confidence.

The Agentic Lending Authority Pack™ ships in three tiers: a fixed-scope adverse-action diagnostic, a full pilot on synthetic applicant data, and an annual enterprise entitlement. Pricing is value-based and disclosed under NDA to qualified applicants.

Canonical KYE™ surfaces referenced on this page: Evidence Pack™ · KYE Protocol™.