AI Receipt Trust Infrastructure
Bankabil AI Trust Fund, powered by Live Oak.
Thesis infographic
Bankabil AI Trust Fund — decision instrument
Recombination equation
Live Oak reserve accounts + nCino workflow evidence + UK model-risk pressure = AI Receipt Trust Infrastructure.
Assessment radar
narrowGating axes (bold) require ≥ 4 to advance from Genesis selection.
Buyer → blocker → evidence
Economic buyer
Blockers
Required evidence
90-day pilot
Can a bank prove, after the fact, what an AI system reviewed, recommended, escalated, and influenced in a lending workflow?
Day 0
Kickoff
Day 30
Buyer discovery + receipt schema v0.
Day 60
Sandbox receipts flowing from one workflow; reserve memo reviewed.
Day 90
Pilot evidence pack; insurer + bank sign-offs; go/no-go.
Success metrics
- ✓95%+ of AI actions produce complete receipts.
- ✓100% of high-risk outputs have human approval or override captured.
- ✓Evidence pack reconstructable in minutes.
- ✓Insurer confirms receipt data is useful for underwriting.
- ✓Bank sponsor confirms reserve model is operationally feasible.
Failure criteria
- ✕Receipts are incomplete or non-standardized.
- ✕Reserve account structure is not approvable by sponsor bank.
- ✕No insurer will engage with the evidence pack.
UK regulatory heat
Decision
narrowAdvance as a Genesis thesis narrowed to AI Trust Account + Receipt Ledger + Assurance Reserve. Defer securitization language.
Next artifact change: Publish Receipt Schema v0 and Reserve Account legal memo.
- Receipt Schema v0 — Product · Week 4
- Reserve legal memo — Counsel · Week 6
- Pilot buyer LOI — GTM · Week 8
One-sentence thesis
Live Oak can make AI risk bankable by combining AI receipts, reserve accounts, insurance wrappers, and bank-grade servicing into a new trust infrastructure for regulated AI use.
Analogy
A mortgage-backed security does not begin as a security. It begins as a mortgage file. The AI equivalent begins with an AI receipt file: model, data, action, approval, policy, exposure, reserve, insurance, and proof. The first product is not "AI MBS." The first product is the AI mortgage file.
Based on the Live Oak Labs Cohort 0 programme, this idea should enter the Lab as a Genesis Studio thesis, not as a finished securitization product. The programme is built to turn Live Oak's founder-lineage primitives into UK-ready bank-infrastructure startups, with every week producing buyer, regulatory, diligence, or pilot evidence — not startup theatre.
Cohort fit — lineage recombination
| Primitive | Translation |
|---|---|
| Live Oak Bank | Small-business banking, reserve accounts, bank-as-lab proof, customer distribution. |
| nCino | Bank workflow standardization for AI governance and evidence trails. |
| Finxact | Ledger / account infrastructure for AI reserves, fees, and trust balances. |
| Payrailz | Payment execution and cashflow movement around claims, reserves, premiums. |
| Apiture | Digital banking interface for AI trust accounts and receipt dashboards. |
| Canapi | Capital formation, strategic bank distribution, venture-scale packaging. |
Genesis scorecard
| Dimension | Score | Rationale |
|---|---|---|
| Buyer urgency | 4 | Banks, SME lenders, insurers, fintech platforms, and regulated AI users need proof of what AI did. |
| Lineage fit | 5 | Bank infrastructure, workflow infrastructure, ledger infrastructure, and capital formation combined. |
| Regulatory tractability | 3 → 4 | Receipt ledger is tractable; insurance and securities layers require licensed partners and careful language. |
| Data feasibility | 4 | Pilot can start with synthetic or permissioned workflow data, then move into real bank-approved data. |
| Pilot clarity | 5 | 90-day pilot can test one AI workflow, one receipt schema, one reserve model, one evidence pack. |
| Integration realism | 4 | Start outside the core; integrate through workflow/API exports before deep bank integration. |
| Capital logic | 4 | Start as SaaS + trust-account fees + insurance partnership; later package pooled receivables / private certificates. |
| Defensibility | 5 | Defensible if Live Oak becomes the bank operating layer for AI accountability. |
Verdict. Survives as a Genesis thesis — but only if the first product is narrowed to AI Trust Account + AI Receipt Ledger + Assurance Reserve, with securitization explicitly pushed to Phase 3.
Final product packaging
Bank product
Live Oak AI Trust Account. For businesses and fintechs using AI in regulated or financially consequential workflows.
Evidence product
AI Receipt Ledger. The standardized receipt layer that proves what AI did.
Capital product
AI Receipt-Backed Trust Certificates. Later-stage institutional product backed by pooled cashflows, reserves, insurance economics, and verified receipt data.
Do not sell "AI securitization." Sell the bank account and receipt infrastructure that could make AI securitization possible later.The AI receipt is the new mortgage file. The AI trust account is the new escrow/reserve structure. The bank is the new servicer of machine-risk obligations.