FIDA: EU Financial Data Access Regulation & Compliance

FiDA stands for Financial Data Access. The European Commission's proposal aims to extend open banking into open finance: financial institutions would make customer data available to authorised third parties through standardised interfaces, at the customer's request. Unlike PSD2 it is not limited to payment accounts but is intended to cover loans, insurance, securities and pensions. Important: FiDA is still a proposal and is not yet in force.

As of August 2026 FiDA has neither been adopted nor entered into force, and there is no binding application date. The sequence so far: the European Commission tabled the proposal on 28 June 2023 (procedure 2023/0205 COD). The Council agreed its position in December 2024 and Parliament decided in December 2024 to enter interinstitutional negotiations. Trilogue talks opened in 2025 and the last round took place in June 2025. Negotiations have been dormant since then, with no further round scheduled as of spring 2026. FiDA remains listed as a pending proposal in the Commission's 2026 work programme, and the withdrawal floated in early 2025 did not happen. Once adopted, a transition period of 18 to 24 months is expected. Plan with scenarios rather than a fixed deadline.

As data holders it would cover banks, insurers, investment firms, fund managers, lenders and other financial service providers in the EU. A notable change from PSD2 is that insurers are included for the first time. On the other side sit data users: regulated financial firms and newly registered financial information service providers (FISPs), which would require authorisation. Whether and how large digital platforms may act as data users is explicitly still open in the trilogue.

PSD2 is limited to payment accounts and to payment initiation and account information services. FiDA reaches considerably further and is intended to cover all material financial data, including insurance, investments, pension contracts and mortgages. The second difference is structural: FiDA foresees that market participants set the technical and commercial rules for data sharing in joint arrangements, the Financial Data Sharing Schemes. PSD2 has no such construct, which in practice produced widely differing interfaces.

An FDSS is a joint rulebook agreed between data holders and data users setting out how data access works technically and commercially: interface standards, liability, dispute resolution and the fees a data holder may charge. The intent is to avoid the fragmentation that characterised open banking. How binding these schemes will be is considered one of the decisive open questions of the procedure.

The proposal names credit and loan data, savings and investment products, mortgages, insurance products and occupational as well as personal pension data, among others. Scope is the central point of contention in the trilogue: the open questions are which data categories deliver value without disproportionate implementation burden, how far back data history must reach, and whether derived or enriched data are included. A final catalogue is therefore not yet settled.

Under the proposal institutions would need standardised real-time interfaces for data exchange, robust consent management with a dashboard where customers grant and revoke permissions, strong authentication, state-of-the-art data security, audit-proof logging of every access and reporting lines to supervisors. Experience suggests the effort sits less in the interface itself than in data quality and in making scattered legacy data fit to be shared at all.

Do not commit a major programme to an uncertain deadline, but complete the groundwork that pays off regardless of the final text. Three items matter most: an inventory of which customer data sits where and at what quality, an assessment of your own interface and consent capability, and a view on whether your institution sees FiDA mainly as an obligation or also as a business model on the data user side. All three retain their value even if the timetable slips further.