Continuous covenant underwriting. The first product running on the schema, and the way insurers price the autonomous frontier first.
What you buy: a continuous, first-party evidence feed on an insured's AI estate, scored against the open covenant schema and delivered to the underwriting desk. Not a questionnaire and not a consulting report; an evidence service the underwriter prices against.
Arkaya supplies continuous, first-party evidence of an insured's AI estate, measured against an eligibility floor at inception and monitored across the term. It replaces the underwriting questionnaire with a record read against obligation rather than self-reported periodically. The underwriter prices it; Arkaya does not.
The covenant schema is class-agnostic. AI liability is the first class written against it, because the underwriting asymmetry is largest there. Cyber, technology E&O, D&O and operational resilience follow.
The governance premium is what you earn by making it visible.
Mainstream carriers are excluding it. The ISO general-liability AI exclusion endorsements took effect on 1 January 2026 across forms underpinning most US commercial liability, and the major carriers have filed to adopt them. A thin specialty cohort writes affirmative cover at the top, a handful of products worldwide.
Exclusion at the bottom and scarce capacity at the top is the wedge. There is demand for first-party and third-party AI cover and no credible basis for writing it profitably without continuous evidence of the insured's AI estate. That evidence is what makes the risk selectable in between.
A machine-readable schema is only as good as the evidence that feeds it. Evidence produced at human cadence, the periodic questionnaire and the point-in-time attestation, cannot describe an estate that changes between checkpoints. Machine-readable underwriting needs evidence produced at machine cadence: captured continuously, at source, as the estate moves. The decision stays human and is recorded; the observation runs at the speed of the thing observed.
Machine-readable underwriting requires machine-cadence evidence. The decision stays human; the observation does not.
A premium is three things: the expected loss, the loading for uncertainty, and the cost of the capital held behind the policy. Evidence acts on the second and third. The uncertainty loading is the documented surcharge underwriters apply to risks they cannot characterise; reduce the ambiguity and the loading falls. At Lloyd's the capital channel flows in full: every syndicate's capital requirement is calculated on an internal model, so evidence that narrows the distribution reaches the capital held, not only the rate.
It also widens the writable book. An applicant below an honest floor gets a route to becoming insurable rather than excluded.
Arkaya supplies the evidence; the underwriter prices it. The calibration layer states how to read a governance evidence record into rating, retention, limit, recovery, reserving and capital, in direction and mechanism. The thresholds and the rate response are the carrier's. The schema and the conformance test are Arkaya's; the calibration, the position and the rating are not.
Every answer carries a grade, and the grade caps the credit available beneath it.
Two insureds, same sector, same limit sought. One produces continuous, verifiable evidence of its AI estate; the other answers a periodic questionnaire. The underwriter reads the difference and prices it.
Illustrative, to show the direction of the read. The carrier sets the rating, the limit and the retention against its own book; these are not Arkaya figures, not a quoted outcome, and not a promise of cover. Direction and mechanism only.
The claim is falsifiable and held as a hypothesis: an evidenced counterparty obtains measurably better terms than an otherwise identical unevidenced one. The result is pending. The experiment is a carrier pilot, and the invitation is open to the desks that can run it. If the evidence does not move the terms, the thesis fails on its own stated ground.
The deliverable is the evidence feed and the record read at the desk: first-party evidence of the AI estate, measured against an eligibility floor at inception and monitored across the term.
The record is the insured's. Portability is a schema property: evidence generated against the schema moves with the insured across engines, insurers and capital events without re-basing. No vendor, and not Arkaya, holds it captive.
The pilot is the §05 experiment run on a live desk: the floor at inception, the term monitored, the terms offered compared against the unevidenced book. If the evidence does not move the terms, the thesis fails on its own stated ground.
Detailed scope, implementation timelines and pricing sit in the post-NDA package: working session, sprint map, term sheet.
The product is one; the buyer is three. Each reads the same evidence and prices a different decision.
Underwriters, syndicates, brokers and reinsurers. D&O, cyber and W&I read against schema-conformant evidence at renewal, claim and exit. Affirmative AI cover and silent-exposure closure in scope.
Acquirers, refinancing lenders, PE sponsors and diligence advisers. Evidence read against debt covenants and SPA triggers, covenant surveillance through the holding period, pre-exit readiness.
Boards, audit committees and risk functions. A Dynamic Risk Maturity Assessment and event-readiness work calibrated to the obligations your next capital moment will test, built to be read by insurers, lenders and acquirers. Detailed scope held under NDA.
One buyer can hold more than one role: an insurer is a subject to its group board, a supervisor of its own coverholders, and a pricer at its desk.
So the reader does not have to infer the distinction.
The factory mutual system has conditioned premium on engineering-surveyed controls since 1835. Telematics repriced motor in both directions. Cyber desks already buy external ratings, whose opaque scores the no-score discipline refuses. Continuous governance evidence extends a 190-year habit; it does not ask the market to form a new one.
The single-domain AI governance platforms each optimise within the AI discipline at the cadence they inherited. None runs every governance discipline at the velocity the environment now sets.
The discipline that calms a sophisticated reader is the explicit statement of what the architecture does not claim. Each line is enforced structurally, not asserted.
Not every exposure is transferred. The deductible and coinsurance every programme keeps is retained by design.
The same evidence that keeps transfer available lets the share retained by design be financed as contingent capital, rather than absorbed by the balance sheet and left unprovided until it crystallises. The instrument heritage is XS Reserve, the excess-of-loss trade credit reserve that won a 2016 Business Insurance Innovation Award, now held within Centinel 10 Ltd.
Resilience Capital is built.
Not asserted.