We did not pivot to AI, we were built for it.
Fable Data’s founder and CEO Luke Kennedy on how AI shapes our business, why it matters, and where it began.
Everyday someone asks me how AI is impacting our business and market.
The answer is simple.
Fable was started 7 years ago. Before ChatGPT, before Claude.
But AI is the age we were built for.
We thought the future lay in machine intelligence, and that those machines would need data.
And the data we thought most interesting is anonymised spending data.
Until now consumer spending data was derived, estimated, modelled, or sampled from surveys published weeks after the fact. Ours is sourced directly from banks. When a consumer buys groceries, books a flight, or pays a utility bill, that event is captured, not inferred.
Aggregated and anonymised across financial institutions globally, the result is a unique picture of the consumer economy broad in geography, granular in detail, and updated every day.
While the organisations we work with are human-driven; governments, central banks, institutional investors, consumer technology businesses and retailers. The interface between our data and their operations is automated.
We think of ourselves as supplying a raw material to AI-powered systems, and that shapes everything about our business.
Daily data was not a feature we added later, it was the architecture we designed around from the start.
A central bank tracking inflationary pressure, a investor positioning ahead of a macro data release, a retailer watching category-level spending shift, none of them can afford to wait for a quarterly estimate. Neither, increasingly, can the AI systems they are deploying to act on these signals.
Anonymisation is not something we do at the end of our process. It happens at the point of collection built into the technology itself, before the data moves anywhere. By the time a transaction enters our pipeline, no individual is identifiable. There is no personal data to lose, expose, or misuse, because our technology ensures it never exists in our systems in the first place.
This makes us fully GDPR compliant, but we try not to lead with that. Compliance is a floor, and we would rather be judged by something more considered than whether we have met a regulatory minimum. We think about this as a question of data ethics, how data is collected matters as much as what it reveals, and we want to be on the right side of that important question rather than just the right side of regulation.
In practice, this matters to everyone in the chain;
For the banks that source data to us, it means they can participate knowing their customers’ privacy is structurally protected, not just promised in a contract.
For the governments and institutions that consume our data, it means they are building on a foundation they can defend.
As AI systems face growing scrutiny over the data that feeds them, we think that foundation becomes more valuable, not less.
We use AI throughout our own processes, anonymisation, collection, validation, structuring, delivery. Not as a novelty, and not primarily as a cost measure, but because it makes the output more consistent and more precise. Which matters rather a lot when the thing consuming your output is itself a machine.
The bar for data quality has shifted in ways that are easy to underestimate. A human analyst brings judgment, context, and a degree of tolerance for the occasional anomaly, whereas an AI model brings none of those things; it amplifies what it receives. Clean data produces sharper outputs. Messy data produces confident errors. We have built accordingly.
The question we come back to is a simple one: what does a machine need from consumer spending data, and how do we deliver it as well as we can? Bank-sourced, globally aggregated, harmonised and anonymised at source, historically deep, and updated daily- that is our answer, for now at least.