Linxo and the Real Cost of Building Open Banking in France
An interview with Gildas Le Louarn, by Sean Murphy
Linxo was connecting French bank accounts long before any regulation asked it to. Founded in 2011, two years after the launch of the iPhone, the company built what its product and marketing director Gildas Le Louarn describes as the first banking application in France: an app that used screen scraping to pull account information from different banks into one place, at a time when the banks themselves had nothing comparable on mobile. Its earliest B2B customers were banks asking for white-label versions of the app.
That first business had a natural expiry date. Once banks built their own applications, Linxo shifted to providing its underlying technology through APIs. Today the company sells account aggregation, data enrichment and payment initiation to banks, credit institutions and fintechs. The consumer app remains, but the two brands that once separated the businesses, Linxo for consumers and Linxo Connect for B2B, have been merged.
The company is today a wholly owned subsidiary of a major banking group, though Gildas is keen to stress that a significant share of its business sits outside the group, which is why it retains a distinct identity. He is equally direct about how the sector's independence ended. When PSD2 arrived, the major French aggregators were independent companies.
In his telling, none are independent today: the directive demanded investment on a scale the independents could not sustain, and each has since been absorbed by a bank or a fund. "For aggregators, PSD2 was a lot of investment we couldn't support, and it is why we are now associated with big banks or funds," he says. The regulation written to open banking to new entrants finished by consolidating those entrants into the institutions they were meant to challenge.
The cost he describes was not only financial. "During two or three years, we didn't develop products, we only developed compliance," he says of the migration from scraping onto the banks' new PSD2 interfaces. Those interfaces frequently did not work, and the customer journeys behind them were poor. Gildas says the company documented the failures, taking screenshots of broken authentication flows to France's national competent authority, and he credits that evidence with the regulator's decision to impose app-to-app authentication, the redirect flow now widely used in France.
The PSD2 use case that has grown most for Linxo is credit: lenders use account access to verify affordability and feed credit scoring, a workflow where regulated access matters because confidence in the data is the product. Cash management and cash pooling for businesses have followed. Personal finance management, the category Linxo was founded on, has barely moved. It was mature before PSD2 and looks much the same after it. The pattern is consistent with the broader European experience of adoption arriving late; Open Banking Limited's impact report counted 13.3 million active open banking users in the UK by March 2025, up 40 per cent in a year, growth that materialised the better part of a decade after the rules did.
On regulation, Linxo has moved from complying with the rules to helping write them. Gildas says the company held long interviews with the European Commission during the drafting of PSD3, pressing two concerns from its PSD2 experience: unclear restrictions on payment flows, and the difficulty national authorities faced in forcing banks to fix non-compliant interfaces. He believes both concerns are reflected in the new text, above all in clearer powers to sanction institutions that fall short.
As Europe looks beyond payments towards open finance, and the debate continues over how frameworks such as FIDA should eventually extend data access across savings, investments and insurance, Linxo's position reflects where it believes the costs actually fall. The expensive part of any access regime, Gildas argues, is the data interface; the consent and authentication flow is comparatively cheap, because every institution already operates one for its own customers. So the company argues for standardising the consent layer, and leaving institutions freer to serve the data itself in whatever way performance and security allow.
Under PSD2, the expense sat in the data interface itself. Banks had to build and maintain dedicated APIs for third parties, and firms like Linxo spent years testing them, working around their failures and rebuilding access they already had. FIDA repeating that model across insurers, pension providers and asset managers, many of them far less technically prepared than the banks were, is what Gidlas sees as the risk in any future framework. Standardised consent with flexible delivery, he argues, would spare them from making the same mistake twice.
