By Emma Steele, Partner at Ascension. Originally published on the Ascension Substack, 29 September 2026.
For the past few years, fintech has had to compete with AI for both attention and capital. The sector that once seemed capable of reinventing every corner of financial services is now being described in much less flattering terms: mature, crowded, overfunded and difficult to differentiate.
The numbers certainly support a more cautious mood. Global fintech investment peaked at $102bn in 2021. In 2025, it was $53bn, roughly half that level, according to Innovate Finance. UK fintech raised $3.6bn in 2025, still second globally, but down significantly from its 2021 high. And the direction of travel has not changed: Innovate Finance’s own headline for the first half of 2026 was that global fintech investment slowed further as AI attracted a growing share of venture capital, with the UK, down 5%, actually outperforming the global market. Some counts are harsher still: Tracxn’s data has UK fintech funding down 26% year on year in the first half of 2026, with late stage funding down 45%.
The obvious conclusion is that fintech has had its moment. I think that conclusion gets something important wrong. Capital cycles and technology cycles are not the same thing, and right now they are moving in opposite directions. Some of the most important changes to how money moves, how financial products are distributed and how financial decisions are made are happening at exactly the moment when competition and entry prices at seed are considerably more attractive than they were a few years ago.
After nearly a decade investing in fintech, we have collected a few scars. We have seen good businesses struggle because the infrastructure was not ready, consumer behaviour had not caught up, or the timing was simply wrong.
Timing as opportunity
Fintech’s graveyard is not full of bad ideas. It is full of ideas that arrived too early, and open banking is probably the clearest example.
When the UK switched on open banking in 2018, there was a huge amount of excitement about what it might unlock. In reality, adoption took years. Consumers were asked to navigate clunky consent flows and give third parties access to their bank accounts, often in exchange for products that were not compelling enough to justify the friction. The underlying idea was right. The timing was not. It took more than seven years for open banking to reach 15 million UK users, nearly one in three adults.
What has changed since then is not just the infrastructure. It is consumer behaviour. Millions of people now interact with AI assistants every week. Software increasingly sits between us and the things we want to get done, and we are becoming much more comfortable with the idea that software can read information, make recommendations and, increasingly, act on our behalf. That matters enormously for financial services, because the psychological barrier that held back some of the earliest open banking use cases, the question of why anyone would let software do this for them, is beginning to disappear.
And the infrastructure is catching up too. Visa and Mastercard have both launched frameworks designed to enable AI agents to make payments with delegated authority. OpenAI and Stripe have developed infrastructure for agentic commerce. Investment into agentic payments startups increased significantly in 2025. The important point is not whether agentic commerce becomes the next giant fintech category. It is that technology has changed the conditions under which previously difficult fintech products can work.
That is a pattern we have seen before. One of Ascension’s first exits was Credit Kudos, an open banking business that rebuilt credit profiles around people’s current financial behaviour rather than historical credit information. Apple acquired the company in 2022. The interesting part, in retrospect, was the timing. Open banking adoption was still moving too slowly for Credit Kudos to scale as quickly as the technology deserved. Apple, however, saw value in owning the technology and developing it inside its own ecosystem.
The same idea can be true in reverse. Sometimes the right question is not why did this fintech fail, but what would have to be different for this idea to work now. That is one of the questions we increasingly ask ourselves, because many of the ideas that struggled between 2018 and 2022, delegated permissions, automated switching and proactive money management, are worth looking at again. The rails are better, the consumer psychology is different and, importantly, many investors have moved on.
AI is not just another fintech product. It is rebuilding the bank.
There is another shift happening underneath the consumer facing innovation.
For decades, retail banking relied on a relatively simple economic model. The current account was often the entry point, and revenue came from overdrafts, fees, deposits and the broader relationship with the customer. That model is under pressure. Consumers are more willing to move their money, regulation has changed the economics of certain products, and challenger banks have forced incumbents to compete on experience and pricing. The interesting question is not whether the old model disappears completely. It is what gets built in its place.
We can already see part of the answer in the economics of the leading digital banks. Revolut reported $2.3bn of pre tax profit on roughly $6bn of revenue in 2025. Monzo reported £172.6m of pre tax profit on £1.7bn of revenue. A decade ago, challengers were routinely criticised for having millions of customers but no clear path to profitability. That criticism now looks dated, as subscriptions, interchange, lending, wealth and other products have created a much broader revenue stack.
But for Ascension, the more interesting opportunity is one layer below the consumer brand. It is the infrastructure that allows both challengers and incumbents to build and operate financial products differently. We are seeing companies applying AI to credit and risk decisioning, fraud and compliance, financial services specific models and synthetic data.
And this is where financial services have a particularly interesting problem. Banks have some of the richest behavioural data in the economy, and they are also among the organisations least able to move that data freely into external AI systems. That creates a fascinating infrastructure opportunity.
One of our more recent investments sits in this space, enabling AI models to be trained inside a bank, using synthetic data, without sensitive customer information leaving the institution. The idea is simple, but potentially powerful: bring the AI to the data, rather than the data to the AI. If that works, the opportunity is not to improve one banking product. It is to change how banks build and deploy a whole generation of models, including credit, risk and fraud models, within the constraints of regulation and customer trust. We think this kind of infrastructure could become an important part of the next chapter of banking.
The consumer lens is alpha, not philanthropy
This is probably the part of our fintech thesis that is most distinctive.
Ascension has spent a long time looking at financial services through the lens of the end consumer. That comes partly from the social impact heritage of our funds, including Fair By Design, where the focus was on tackling the poverty premium (the extra cost of being poor). For a long time, that perspective could be interpreted as a values based choice. We increasingly see it as an investment advantage, because in consumer fintech, understanding why someone changes their behaviour is often more valuable than understanding what technology can technically do.
Take Wagestream, now Stream. We backed the company at pre-seed in 2018 through Fair By Design, when earned wage access was still a relatively niche product for frontline workers. The initial product was the wedge, but the deeper insight was about behaviour: if you give people a useful financial service at exactly the point where they need it, you can earn repeated engagement and trust.
Since then, the company has expanded well beyond earned wage access into a broader workplace finance platform spanning savings, budgeting, affordable credit, financial planning and pensions. In January 2026, it raised a $90m Series D and announced further expansion in the US. The company has also been publicly exploring the possibility of an IPO, although the timing and listing venue remain undecided.
For us, that makes the original investment thesis particularly satisfying. The wedge got them in, the consumer insight built the platform, and what looked like a niche product in 2018 has grown into a financial platform supporting millions of workers across thousands of employers. The lesson is not that every fintech should start with earned wage access. It is that the best consumer fintech companies often start with a very specific behavioural insight and then expand from there.
We see something similar in mortgages. A mortgage is one of the biggest financial decisions most households make, and it is also highly emotional. That creates an unusual opportunity: if a product can turn a traditionally episodic financial decision into something genuinely useful in people’s everyday lives, engagement can become a competitive advantage.
Tembo, which we backed at pre-seed alongside Aviva, and has since gone on to raise its Series C, helps first time buyers unlock family support to get onto the property ladder. Sprive, where we led a £5.5m seed round, now post Series A, turns everyday spending into automatic mortgage overpayments. Different products, but a similar underlying principle: find the emotional hook that makes people come back. Acquisition is important, but in consumer fintech, retention and trust are often where the real moat begins.
So what are we looking for?
We do not think every fintech company suddenly becomes attractive because AI exists. Quite the opposite. AI makes some fintech businesses easier to build, which means it also makes some of them easier to copy. So our questions have become more specific.
Where does the AI advantage accrue? If a startup is simply putting a general purpose model in front of an incumbent’s data, the incumbent may ultimately capture much of the value.
What survives technological change? A product needs some combination of distribution, trust, proprietary data, regulation or customer behaviour that cannot simply be replicated by adding an AI layer.
Is there a data advantage that generalist AI companies cannot easily access? Regulated financial data is difficult to move and difficult to use. That friction can be a problem, but it can also create defensibility.
And finally, does the product genuinely improve the customer’s financial position? We do not see this as philanthropy. Products that save people money, make financial decisions easier or improve access to financial services tend to have better retention, and they also tend to be more aligned with the direction of regulation. The consumer outcome and the investment case can be the same thing.
What this means for founders
There is no point pretending that fundraising is easy right now. It is not. Late stage fintech funding has contracted sharply, and founders should plan for rounds to take longer and be smaller than a few years ago.
That makes the fundamentals matter more, not less. Build runway, know your unit economics, be disciplined about where every pound goes, and do not build your company around the assumption that the next round will arrive exactly when you need it.
But there is another side to this. A difficult funding environment can be a very good environment in which to build. Competition is lower, valuations at the earliest stages are more rational, customers are still spending money, the underlying infrastructure is improving rapidly, and some of the technologies that were too early a few years ago may finally be ready.
The founders who build through this period may find themselves in a very different competitive environment when capital becomes abundant again: fewer competitors, better infrastructure, more sophisticated customers, and a market that has finally learned which problems are worth solving.
Fintech being uncool is a gift
The best time to invest in a category is rarely when everyone agrees it is the future. Fintech has spent a decade attracting enormous amounts of capital, talent and attention. Some of that enthusiasm was justified, and some was not. The correction has been painful, particularly for founders who raised at the top of the market, but it has also created something we have not seen for a while: space to think.
We can look again at ideas that failed because the technology was not ready. We can invest in the infrastructure and workflows underneath the next generation of financial products. We can focus on companies that understand consumer behaviour rather than simply adding an AI interface to an existing product. And we can do all of that without having to compete with every generalist investor who decided, five years ago, that fintech was the obvious place to be.
That is why, after nearly a decade investing in the sector, I think this is one of the most interesting moments we have seen. Not because fintech is booming, but because it is not. The tourists have left. The founders building through this period are the ones we want to know, and we intend to back them.
Emma Steele is a Partner at Ascension, investing in fintech, health and climate founders at the earliest stages. If you’d like to stay in touch or speak with the team or thinking of partnering with us (whether through mentorship, expertise, direct or fund investment) feel free to reach out to info@ascension.vc
About Ascension
Ascension has been on the side of early-stage founders for over a decade. With 150+ portfolio businesses and 15 profitable exits, we are conviction-led investors guided by the belief that frontier technology has the ability to solve some of our most pressing global issues.
We back the founders reshaping the fabric of our world.
Across FinTech, Climate and Health, verticalised AI and deep tech are transforming the systems underneath everyday life. We invest at pre-seed, backing the founders with the intent and technical ambition to reshape them.
Fintech has been an intrinsic part of Ascension’s journey and we will continue to focus on future-proofing financial resilience by looking at the whole value chain, whether it is on the consumer side all the way to looking at the infrastructure sitting beneath it.