Two of France’s largest banks suffered separate open-banking connectivity outages through Tink early Sunday, with BNP Paribas recovering only 12 minutes before Crédit Agricole entered its own disruption.
Tink’s automated monitoring detected the BNP Paribas outage beginning at 05:30 CEST on October 4. The provider publicly reported the incident at 05:50 and marked it resolved at 06:32, giving the disruption a duration of approximately 62 minutes.
Crédit Agricole then entered a separate outage beginning at 06:44 CEST. Tink reported that incident at 07:04 and restored the connection at 07:52, meaning it lasted approximately 68 minutes.
The two outages therefore began 74 minutes apart, while only 12 minutes separated the resolution of BNP Paribas from the start of the Crédit Agricole failure.
Tink’s official France provider status page did not identify which flows were affected, what percentage of customers encountered failures or what specific bank-side error occurred. Both incident entries list the affected flows and users as unavailable.
That lack of technical detail means the close timing should not be treated as evidence of a shared failure. There is currently no public indication that BNP Paribas and Crédit Agricole were disrupted by the same certificate problem, PSD2 infrastructure provider, network dependency or software change.
Tink’s Own Platform Did Not Report an Outage
One important piece of evidence argues against a broad Tink platform failure.
Tink’s main service status page recorded no incidents on October 4. Its payments, transactions, account-check and other central services remained listed as operational.
Tink separates the health of its own platform from the connectivity of individual banks in each market. The France page tracks providers including BNP Paribas, Crédit Agricole, Société Générale, Banque Populaire, Boursorama and others.
That architecture matters because a fintech using Tink can experience what appears to be an outage even while Tink’s central API remains healthy. The failure can sit further downstream in the connection to a particular bank.
The distinction resembles the infrastructure problem recently visible in Xendit’s repeated payment failures across Indonesia and the Philippines. Xendit’s central platform remained largely available while separate e-wallet, QR and banking partners experienced disruptions underneath it.
Open banking creates a similar dependency chain. A fintech may integrate with one API provider, but a successful user journey can still depend on the bank’s PSD2 interface, authentication systems, certificates, network connectivity and Tink’s own integration layer all functioning at the same time.
Crédit Agricole Had Just Recovered From a Much Longer Outage
Crédit Agricole’s October 4 incident stands out because it was not the bank’s first recent problem on Tink.
Tink recorded another Crédit Agricole outage beginning at 04:29 CEST on October 2. That incident was not resolved until 08:00 on October 3, producing a disruption of roughly 27 hours and 31 minutes.
The provider page again did not identify the affected flow or percentage of users.
That means Crédit Agricole returned to an outage state less than 24 hours after Tink marked the longer incident resolved.
The pattern does not prove the two Crédit Agricole events had the same root cause. A bank can experience unrelated failures across authentication, account-information and payment-initiation services. But the recurrence makes the October 4 event more significant than an isolated one-hour interruption.
Crédit Agricole had also experienced other interruptions in September, including a September 20 payments outage affecting 52% of users in Tink’s monitoring and a major September 21-22 manual-authentication incident listed as affecting 100% of users.
BNP Paribas Has Also Had Recent Tink Connectivity Problems
BNP Paribas has its own recent history of interruptions in Tink’s French provider network.
On September 24, Tink reported an auto-refresh outage affecting 54% of BNP Paribas users. Earlier, a major incident beginning September 21 affected manual authentication for 100% of monitored users before being resolved the following day.
That does not make Sunday’s outage unusual enough to imply a deeper systemic problem by itself.
Open-banking integrations regularly experience short disruptions because the service path crosses several independently operated systems. What makes October 4 interesting is the sequence: two major French banks stopped functioning through the same aggregator within a tight morning window.
The natural question is whether something common sits underneath both connections.
PSD2 Makes Bank APIs Part of Everyday Financial Infrastructure
Tink launched its French open-banking operations around the PSD2 framework, connecting its platform to APIs provided by major French financial institutions.
Under PSD2, authorized third-party providers can access account information or initiate payments when customers provide consent and complete the bank’s required authentication.
Tink effectively abstracts those separate banking connections behind a common infrastructure layer. That is valuable because fintech companies do not need to build and maintain completely independent integrations with every bank they support.
But abstraction does not eliminate the underlying dependencies.
A similar business logic is driving companies building payment infrastructure that connects multiple local banking rails through a single API. The product becomes easier for developers to consume, while operational complexity moves into the infrastructure provider.
That makes reliability one of the central economic propositions of open banking.
A consumer does not care whether an account refresh failed because of Tink, BNP Paribas, an expired certificate or an authentication endpoint. They simply see that their banking application cannot connect.
The Missing Flow Data Makes the Impact Hard to Measure
Tink’s incident records normally become much more useful when they identify the affected function.
Recent France incidents have specified failures involving payments, authentication, manual refresh and automatic refresh. Some have also shown the share of users affected.
Neither October 4 bank outage contains those details.
That prevents a meaningful estimate of customer impact.
An account-information outage could prevent budgeting apps and financial dashboards from refreshing balances. An authentication problem could stop users from reconnecting their bank accounts. A payment-initiation failure would be more directly transactional, potentially preventing customers from completing bank-to-bank payments.
Without the flow data, it would be misleading to describe either incident specifically as a payments outage.
This is also why outages inside financial infrastructure can be harder to interpret than ordinary website downtime. A service may remain partially functional while one workflow fails, similar to how individual bank-transfer failures can occur inside an otherwise operating financial platform.
The Close Timing Is Worth Investigating, but Not Overinterpreting
The temptation is to look at BNP Paribas failing at 05:30, Crédit Agricole failing at 06:44 and assume there must be one cause.
That would be premature.
Tink classified the incidents separately. Its own central platform did not report a corresponding outage. No public notice currently identifies a common network provider, cloud service, certificate authority or PSD2 technology vendor involved in both events.
There are several plausible explanations.
The banks may simply have experienced independent maintenance or infrastructure failures during a low-traffic Sunday morning window. Tink’s monitoring could also have detected separate compatibility problems after changes made independently by each bank.
Alternatively, there could be a shared dependency that has not yet been publicly disclosed.
Open-banking systems depend heavily on mutual TLS certificates, OAuth authentication, eIDAS credentials, API gateways and networking layers. A common service provider or certificate-related change could theoretically affect several banks without taking Tink itself offline.
But theory is not evidence.
Until another aggregator reports matching failures or one of the banks publishes a technical explanation, the most defensible interpretation is two closely timed provider outages with an unproven relationship.
Repeated Short Outages Expose the Weakness of Aggregated Finance
The bigger story is not necessarily whether Sunday morning’s two failures were connected.
It is how much modern financial software now depends on invisible intermediaries.
A customer may open one fintech app and see accounts from five banks in one place. That feels like a unified financial service.
Underneath, it can involve five separate bank APIs, a third-party provider such as Tink, authentication infrastructure and the fintech’s own systems.
The same fragmentation is visible when payments become delayed between different financial institutions: the customer sees one transfer, while the transaction itself crosses multiple operational layers.
Open banking improves competition partly by hiding that complexity from users.
The downside is that failures also become harder to locate.
Tink’s segmented status pages are useful precisely because they expose whether the problem sits with the central platform or a specific bank connection. On October 4, that distinction points toward the bank-integration layer rather than a Europe-wide Tink failure.
What Would Turn the Timing Into a Bigger Story
The most useful next evidence would come from another open-banking provider.
If a separate aggregator recorded BNP Paribas and Crédit Agricole failures during the same windows, that would strengthen the argument that the incidents originated on the banks’ PSD2 infrastructure rather than inside Tink-specific integrations.
A disclosed certificate rotation, API version change or common network provider would make the simultaneous timing more significant still.
For Crédit Agricole, recurrence is the other signal to watch. A second disruption less than a day after a 27-hour outage raises the question of whether the earlier restoration addressed the root cause or only restored service temporarily.
For now, both banks are operational again on Tink.
The confirmed facts remain narrower: BNP Paribas was unavailable through Tink for about 62 minutes, Crédit Agricole entered a separate 68-minute outage shortly afterward, Tink’s central services stayed operational, and the public incident records do not contain enough technical information to connect the two events.
The 12-minute gap between them makes the coincidence worth investigating.
It does not yet make it a common outage.
Johan Shamshad is a financial markets writer at Dave Finances covering cryptocurrencies, trading platforms, brokers, fintech, financial regulation, and developments across global markets. He previously worked at Gulf News, adding newsroom experience to his coverage of fast-moving financial and digital-asset markets.
His work focuses on identifying market-moving events, company developments, regulatory changes, product launches, and shifts in trading and financial infrastructure.
Johan contributes news and analysis designed to help readers understand not only what happened, but why a development matters and how it may affect the wider financial landscape.