The question of ‘who should pay’ for open banking has been a feature of the ecosystem for several years and has recently become a major talking point in the US.

Back in July, JPMorgan hit the headlines with the news it intended to secure fees from open banking infrastructure providers (such as Plaid, Akoya, Mastercard and MX) for accessing account data through its APIs. The bank argued that it should be compensated for the costs of providing this access, particularly as volumes grow.

This wasn’t the only development in this area. The JPMorgan announcement was followed in late August by the news that the CFPB will be exploring options for cost recovery models as part of its consultation into potential changes to the Personal Financial Data Rights rule (PFDR). While this will go through the consultation process, it does put this issue very much back on the table as any revisions or updates to the current PFDR are made.

In the past week, it was announced that JP Morgan and Plaid had reached an agreement over the fee structure for the latter to compensate the bank for access to its APIs.

While charging for API access to customer account and transaction data may seem like the proverbial ‘no brainer’, this is an area in which banks should tread extremely carefully.

There are three reasons.

Open banking APIs are a channel, and it’s your customers that are driving demand

Banks should remember that the ultimate driver of growth in API call volumes is not open banking infrastructure providers or fintechs, but your customers wanting to use their account data in third party apps and experiences. Viewed this way, open banking is more akin to a new digital channel and should be considered as such.

When considering fee models, a careful balance needs to be struck to ensure that any income doesn’t come at the expense of the downstream customer experience. Pitching too high will deter API use after all, which will have a potential impact on the services customers use. And if someone using their favourite fintech app finds they can’t view data from a particular bank anymore, or it starts to work less well, or they face fees to do so, there’s a good chance the bank will take some of the blame.

In this scenario, the costs of managing a disgruntled customer may come to far outweigh the fee income from API calls.

Excessive fees create the business case for API workarounds

It’s also important to remember how we got here.

A big part of the business case for building external facing APIs for account data was to replace screen scraping and other unofficial mechanisms for accessing this information. The rationale being that migrating that activity to a series of APIs brings far greater security and puts banks in control of the data that is shared. This also has the benefit of reducing peak transaction volumes on the channel infrastructure.

Any plans around fee structures should therefore consider the risks of pushing the market back to screen scraping and similar alternatives. As well as reintroducing risk to the bank and inconvenience to customers, it would potentially drive higher costs in other areas of the bank (should traffic move back to the web interface for example).

Digital banking agents will come to rely on open banking APIs

Agentic AI is one of the hottest topics in the industry today, and with good reason. While the idea of a customer’s digital agent interacting autonomously with the bank seems like something from the distant future, the window for action is narrowing and banks should be considering how they will support digital agents as part of their future channel offering.

The functionality offered by today’s open banking APIs makes them an ideal underpinning for many agent-led models or use cases. For a bank wishing to play in this space then, both having this infrastructure in place (and potentially the ability to access the APIs of other institutions) would become critically important.

This brings things full circle on the ‘open banking as a channel’ point. It’s not that many years since some argued that charging for mobile banking access was a viable and sustainable model, and we know how that played out.

Time will tell if the same logic holds for open banking.