Agentic Credit Broking ProtocolSeamless credit broking journeys delivered through user-controlled AI assistants, agents and applications
The ACBP is a new, global, and open protocol that preserves regulatory accountability, and allows AI providers, credit brokers, and lenders to deliver best-in-class AI experiences
View ACBPView ACB ProtocolThe problem
View ACBPView ACB Protocol
Users start in AI, but can't stay there
AI assistants have no way to participate in regulated credit broking. The user must be redirected to the broker's own interface.
Mediation and regulation are coupled
Who mediates the interaction and who carries regulatory responsibility are treated as the same thing. Only the broker's own surface can conduct the journey.
No trust model for agents
There is no fixed code path to audit. Trust must come from observing behaviour, not inspecting source.
The model
View ACBPView ACB ProtocolAll user agents can open a case and broker credit
The User Agent pushes financial data to the broker and reads back structured events using the protocols. Any agent that can call MCP tools can participate.
Broker controls compliance
Disclosures, consents and declarations are all handled through the protocols ensuring that the process steps and interactions are handled compliantly.
Trust is progressive and efficient
Unknown user agents default to the lowest-trust path. This ensures safety by default with trust being earnt through the process.
Why it matters
View ACBPView ACB ProtocolUsers stay in their chosen assistant throughout the journey.
Brokers distribute through every AI channel without losing regulatory control.
Lenders gain new distribution channel whilst trusting lending is appropriate and enforceable.
Regulators gain structured, replayable audit trails for every journey.