Fintech & Banking

Open banking in Australia: what the Consumer Data Right delivered

Open banking in Australia: what the Consumer Data Right delivered

7 min read·1570 words·Updated 22 September 2026

In 2019 I was the product lead at a payments start-up, sitting in a pitch meeting where our founder described the Consumer Data Right as “the biggest thing to happen to Australian banking since the four pillars”, and the investor across from us wrote a number on a pad and underlined it twice. We got the money. We built the switching product. Nobody used it, and the company I’m describing is one of the two on my résumé that didn’t make it. I’ve thought about that meeting a lot, because the CDR didn’t fail, exactly. It did something quieter and more useful than the pitch, and the companies that noticed are the ones that are still here.

What it was supposed to do #

The Consumer Data Right is a legislated framework, administered jointly by the ACCC and the privacy commissioner with Treasury setting the rules, that gives you the right to direct a business holding your data to share it with an accredited third party. Banking was the first sector, live from July 2020 for the big four and progressively for the rest. Energy followed. The idea was competition: if you could hand your transaction history to a new lender in thirty seconds, switching would get easy, and easy switching would discipline the incumbents on price.

The architecture was, and is, genuinely good. Standardised APIs, a consent model that’s stricter than almost anything overseas, a public register of accredited recipients you can check on the CDR site. Australia built the plumbing properly. What it didn’t build was a reason for most people to turn the tap.

Books Reading — Consumer Data Right
Photo: Raymond Sam via StockSnap (CC0 1.0)

What actually happened #

Consumer uptake was small and stayed small. The published participation figures, which the ACCC releases, showed consents in the hundreds of thousands at a time when the population of bank customers is in the tens of millions. The reasons are not mysterious. Switching banks in Australia was already possible and most people didn’t; the barrier was inertia, not data. The consent flow, designed to be safe, was also long. And the recipients who could use the data were mostly fintechs that the public had never heard of, asking for access to everything.

Meanwhile the cost of becoming an accredited data recipient ran, for a small company, into the hundreds of thousands of dollars in compliance, security assessment and insurance, before any product was built. A lot of the early cohort spent their seed round on accreditation. The intermediary model, where a handful of accredited platforms collect the data and pass it through to unaccredited businesses under stricter rules, eventually eased that, and it’s the model most survivors now use.

There was also a design choice that looked prudent and turned out to be a brake. Consents expire, by rule, after a maximum period, and the customer has to renew them. For a lender making a one-off decision that’s fine. For an ongoing product, a budgeting app or a cash-flow monitor, it meant asking the customer to re-authorise through the full bank flow every year, and a meaningful share of them simply didn’t. The businesses that depended on a persistent connection to a customer’s accounts, which was most of the consumer-facing cohort, watched their active user numbers reset with the consents. It was a privacy protection doing exactly what it was designed to do, and it’s a good illustration of why the scheme worked for lenders and not for apps.

The reset #

By 2024 Treasury had heard enough. The government announced what it called a reset of the CDR that August: a narrowing of focus back to banking, a pause on expanding into new sectors, a concentration on the use cases that were actually working, and a commitment to lower the compliance cost for the businesses that had to build the data-holder side, which was, by then, most of the credit unions and small banks in the country complaining about the bill. The consultations since have been about making the consent flow shorter and the data more useful to lenders in particular. Action initiation, the ability to instruct a payment or open an account through the CDR rather than just read data, was legislated in 2022 and has spent the years since waiting to be switched on.

I’d call the reset an admission, and a sensible one. The scheme was built as a universal right and adopted as a niche product, and the government stopped pretending otherwise.

How a Consumer Data Right share works. And where the value ended up
Diagram: Technology Spectator

What it quietly delivered #

Here’s the part the founder in 2019 got right without knowing it. The CDR made lending faster. A lender that can pull twelve months of verified transaction data with consent doesn’t need to ask for bank statements, doesn’t need to worry about doctored PDFs, and can make a responsible lending assessment in minutes rather than days. That’s why the businesses that survived are, disproportionately, in credit: mortgage brokers’ tools, personal lenders, small-business lenders, the lending arms of the buy-now-pay-later companies once regulation pushed them toward proper assessment. Nobody switched banks because of the CDR. A lot of people got a loan decision on the same day because of it, and the industry stopped emailing PDFs of bank statements around, which I’d count as a public good in itself.

The second thing it delivered was a standard. Before the CDR every fintech that wanted your bank data used screen-scraping: you handed over your internet banking password and a bot logged in as you. The CDR gave the industry a legitimate alternative, and while scraping hasn’t been banned, its days are visibly numbered. That was worth doing on its own.

A third, smaller delivery: fraud. A lender that receives transaction data through the CDR’s authenticated channel knows the data came from the bank, not from a customer’s edited spreadsheet. The mortgage broking industry had a persistent problem with altered bank statements, documented in ASIC’s reviews of the sector, and the verified channel closed it for the lenders that adopted it. Nobody wrote a press release about that either.

Where the fintechs went #

The founders I talk to now describe the CDR as infrastructure, in the way a road is infrastructure. Nobody raises money on the road; they raise money on what the road lets them deliver. The pitch decks stopped saying “open banking” around 2022 and started saying “cash-flow underwriting” or “verified income” or “real-time affordability”, which are the products. A friend’s company from the same cohort made exactly that pivot, from a consumer switching app that nobody used to a lending decision engine that several lenders now pay for. We didn’t, which is the difference between the two lines on my résumé. Same data. Same plumbing. Different customer.

The energy CDR, for what its worth, has followed the same pattern in miniature: consumer uptake negligible, a small number of comparison and demand-management businesses using it in the background, and a reset conversation of its own.

What the data holders paid #

The side of this story that gets the least coverage is the cost to the businesses that had to make the data available. The big four banks absorbed it; they have technology budgets in the billions and the APIs were, for them, one project among hundreds. The mutuals and regional banks did not have that luxury. Building and maintaining a compliant data-holder platform, with the consent management, the performance standards, the reporting to the ACCC and the ongoing rule changes, cost a small credit union a sum that its members will never see a benefit from, because almost none of them will ever use it. The Customer Owned Banking Association was blunt about this in its submissions, and the reset’s commitment to lower the compliance burden was, in large part, an answer to that sector.

It’s the structural flaw in a universal right: the cost falls on everyone and the benefit accrues to the few who use it. A narrower scheme, built around the lending use case from the start, would have cost the mutuals a fraction and delivered the same outcome. Hindsight, obviously. But it was hindsight several people had in 2018 and said out loud.

The measured opinion #

The mistake, in my view, was selling the CDR to the public as a consumer feature rather than to the industry as a utility. Consumers don’t want a data right; they want a cheaper mortgage and a shorter form. The scheme delivers the second of those and was marketed on the first. The Treasury and ACCC material at the time was full of switching, and switching was never the prize. If the reset lands on that realisation, and the recent Treasury consultation papers suggest it has, the next five years will look better than the first five, and quieter.

I had coffee last week with my old founder from that 2019 meeting. He’s raising again, for a different company, and the deck doesn’t mention the CDR once, even though the product couldn’t exist without it. That’s what infrastructure looks like when it works.

The water at Bronte this morning was 16 degrees and the swim was short. I mention it because it’s the hour of the day when I think most clearly about the things I got wrong, and this column is about one of them: a thing that was supposed to make money made something else instead. I reckon that’s the more common outcome in this industry than anyone’s deck admits.

— Sophie Lindqvist, AI, Start-ups & Fintech Editor, Technology Spectator

Back to top button