03 / Use cases

Where autonomous finance stops today.

The technology already works. What is missing is the ability to prove authority at the moment of interaction.

Scenario 01

SME lending · Challenger bank

A credit agent cannot verify who it is dealing with.

Today

Bank A's credit agent needs to confirm counterparty authority at Bank B before committing to a syndicated facility. No standard method exists, so a human steps in or the deal stalls.

Breaks without IAP

Liability exposure with no record of what was authorised
A manual bottleneck at the point that needed to be fast
Autonomous lending cannot scale past pilot volume

With IAP

Identity, mandate and proofs clear in milliseconds and the facility settles straight through. A failed provenance chain never reaches the counterparty — it lands on a human's desk.

Accountable: Chief Risk Officer / Chief Compliance Officer

Scenario 02

Inter-bank liquidity · Tier 1

Treasury agents spot the window and miss it.

Today

Overnight liquidity opportunities need human sign-off across institutions — a process measured in hours, against a window measured in minutes.

Breaks without IAP

Dead capital sitting idle overnight
Yield forgone on opportunities correctly identified
The human bottleneck worsens as agent coverage grows

With IAP

In-window positions settle at machine speed with limits enforced in hardware. Treasury sees only exceptions, with the failing check named and evidence attached.

Accountable: Head of Treasury / Chief Risk Officer

Scenario 03

Regulatory audit · EU AI Act

The regulator asks why, and the answer is a black box.

Today

An AI agent makes AML decisions. The regulator asks who authorised it, what its limits were and why it flagged a transaction. The firm cannot produce a tamper-proof answer.

Breaks without IAP

No tamper-evident record of authority or constraint
Black-box liability sitting with a named individual
Exposure to enforcement action and remediation cost

With IAP

Every interaction writes a signed manifest and a hash-chained ledger entry — showing what the agent did and which controls stopped it doing anything else.

Accountable: Chief Risk Officer / Chief Compliance Officer

03.4Market expansion

One control layer, every negotiated market.

The eight checks do not change when the instrument changes. Each new market is a rulebook and a counterparty onboarding, not new protocol engineering.

Market	What agents negotiate	Scale of prize	Agent readiness	Phase
Interbank lending	Tenor, rate and counterparty exposure inside a credit envelope	Very large	Live today	Entry
Money markets	Overnight and short-dated placements against a daily liquidity mandate	Very large	Live today	Entry
FX	Price, size and settlement window inside a per-pair risk limit	Very large	Live today	Entry
Repo	Rate, tenor and collateral schedule against an eligibility rulebook	Very large	Emerging	Expansion
Collateral optimisation	Substitution and allocation across a haircut and concentration policy	Large	Emerging	Expansion
Securities lending	Borrow fee, recall terms and collateral basket per lending mandate	Large	Emerging	Expansion
Firm payments	Routing, value date and beneficiary checks inside a sanctions and limit policy	Very large	Emerging	Expansion
Bonds	Price, size and settlement against issuer and duration limits	Very large	Early	Horizon
Equities	Block terms and execution constraints inside a best-execution policy	Large	Early	Horizon
Derivatives	Structure, margin terms and novation inside an approved product list	Very large	Early	Horizon
Trade finance	Document conditions, tenor and pricing against a country and sector policy	Medium	Early	Horizon

Scale of prize is a relative ranking of institutional flow across these markets, not a revenue forecast or a claim on addressable market. Agent readiness reflects how far automated execution has already progressed in each market today.

03.5Go to market

Starting with UK challenger banks.

A deliberately narrow entry point: modern stacks, fewer approval layers, and agents already live in SME lending and liquidity.

Shorter sales cycles

Fewer approval layers between the risk owner and the deployment decision.

Direct feedback loops

The people running the pilot are the people accountable for its outcome.

Referenceable proof

A regulated deployment that tier-one institutions can evaluate before committing.

Discuss a design partnership