On 8 and 9 June 2026, the FCA's Smart Data Accelerator brought around 80 stakeholders - banks, building societies, fintechs, third-party providers, credit reference agencies, trade bodies and consumer groups - into its offices to work through what open finance would actually need to improve mortgage journeys. The outcomes report, published 3 September 2026, states access to more data is not, by itself, the main challenge. Success depends on whether that data can be treated as trusted evidence, and whether the infrastructure, accountability and participation arrangements needed to support it are in place.
That's the right diagnosis. But read the report closely, and every mechanism it proposes for building trust is aimed at the data itself. None of it is aimed at the entities moving that data.
Participants coined a useful term: "decision-grade data" - information a lender can actually rely on because it is "accurate, current, standardised and capable of being recognised as evidence." To get there, the report says participants "repeatedly called for common data models, minimum requirements, data provenance, auditability and trust marks."
Every one of those proposals describes a property of the data. None of them describes a property of the organisation supplying, transforming, or holding it. A trust mark on a dataset says something was accurate when it was captured. It says nothing about whether the aggregator, wallet provider or technology platform that captured, transformed, and is still holding it today remains the same reliable counterparty it was when the mark was issued.
The report's clearest articulation of its future-state vision is a phrase several teams landed on independently: "capture once, verify once, reuse safely across the ecosystem." The report defines this as a model where "consumers provide information once, that information is verified at source or through trusted infrastructure, and it can then be reused for appropriate purposes with the consumer's control."
Notice what gets verified once: the information. Notice what doesn't get mentioned at all: whether the infrastructure doing the verifying, or the entity holding the reusable copy afterwards, is checked again at any point after that first verification. A mortgage journey can span years - readiness, application, in-life management, later-life planning, exactly the four stages the sprint's own personas were built around. A data point verified once at the start of that relationship says nothing about whether the aggregator or technology provider still handling it three years later is the same risk it was on day one.
The report is specific about who sits in an open finance-enabled mortgage journey: "data holders, wallet providers, aggregators, brokers, advisers, lenders, public bodies, technology providers and AI tools." That's nine distinct categories of organisation, each with different access, different roles, and by the report's own account, no common framework yet for how any of them - beyond the data they individually handle - gets assessed, monitored or re-assessed over time.
Figure 6 in the report lists seven "foundational components" the ecosystem needs: data standards, interoperable interfaces, a consent framework, digital verification, a governance model, commercial arrangements, and safeguards. "Digital verification" is defined narrowly - confirming "that data belongs to the consumer" - and "governance model" stretches only to defining "participant roles, standards, performance expectations, liability and redress." Roles and standards are set once. Nothing in the list describes checking, on a standing basis, whether a participant is still meeting them.
The report's final paragraph is key. The Smart Data Accelerator's next phase of work, it says, will look at "the data-sharing architecture, roles and responsibilities, data journeys, the key components of infrastructure and trust such as identity and verification, authorisation and certification, and the interdependencies between agentic AI and open finance."
Identity and verification. Authorisation and certification. That is entity-level trust infrastructure, named as the FCA's own next research priority - the regulator's own stated direction. The gap this piece has described isn't a criticism of the sprint. It's the item already at the top of the FCA's own list for what comes next.
-p-2000.jpg)
