How AI Will Transform Fintech in 2026
The core of this conversation is not simply about how 'AI will make things more convenient.'
It is much bigger. The very subject of the word 'fintech' is changing.
In the past, fintech meant 'startups doing bank-like things outside of banks.'
However, the speakers now state clearly: fintech has become synonymous with financial services.
In other words, the 'movement to rebuild finance with software' is no longer a peripheral movement; it has moved into the center of the financial industry.
With that in mind, what is the focus for 2026?
It is that (1) improving the quality of finance, (2) AI taking over financial 'tasks,' and (3) as a side effect, a massive surge in fraud will all happen simultaneously.
And, importantly, the speakers are not talking about science fiction. On the contrary, they are describing an extremely realistic change where AI dismantles the 'mountain of manual work' remaining in financial operations.
1. Fintech: Reading 2018–2025 through 'Investment and Product'
1-1. 2018–2019: 'The Slow Spring' = The era when the industry 'got a name'
2018–2019 was the period when fintech was recognized as having 'established itself as an industry.'
The changes during this period were closer to a redesign of the consumer experience than to technological progress.
Securities: Opening an account via smartphone → immediate trading
Banking: No need to visit a branch → completed via app
Remittance: Crossing borders → simplified digitally
Crypto assets: Speculation, exchange, and holding became 'app-ified'
A symbolic example cited is an entity like Robinhood.
The UI shattered the common wisdom that 'investing is difficult/has a high barrier to entry.'
The startups born at this time cut out financial functions as 'components' and reassembled them in places closer to the user.
The main theme of this stage would later be clarified: 'solving the access problem.'
In short, they made it possible to do things on a smartphone that previously required visiting a bank branch. That alone created enormous value.
1-2. 2020–2021: 'The Roaring Summer' = The era when capital 'flooded into fintech'
What came next was the 'summer' brought on by COVID-19.
The strongest figures in the discussion appear here.
'Approximately 25% of venture dollars during this period went into fintech'.
What happened here was that changes on the demand side and changes on the capital side meshed simultaneously.
Demand side: People went out less, life moved online, and digital finance became essential
Capital side: A low-interest-rate environment persisted, and funds concentrated in growth stocks and high-growth companies
And fintech was powerful as a 'narrative.'
'Reinventing banks,' 'zero fees,' 'liberating the individual.'
Such rhetoric resonated with both investors and consumers.
However, from the perspective of those involved, there was another side to it.
The speaker compares that time to 'EDM starting to blast at full volume.'
In other words, while the frenzy was fun, it was also an environment where making correct decisions became difficult. When growth is too fast, it becomes hard to see the quality of products or the differences in organizational strength. Because everyone is going up.
1-3. 2022–2024: 'Winter' = The era when capital withdrew and businesses were tested
And then, from the latter half of 2022, a "winter" arrives.
What is being discussed is a somewhat ironic contrast.
"After 25%, it's almost 0%".
The important point here is that the cause of the winter was not the "end of a trend," but a structural change in interest rates and the cost of capital.
Low interest rates: Models based on borrowing and expanding are easy to sustain (loan margins are generated, growth is rewarded with a premium)
High interest rates: The cost of capital rises, putting pressure on lending models (margins shrink, regulations also become heavier)
What happened during this winter?
Simply put, it was a "culling," but more accurately, it was a redefinition of companies.
Neobanks: You cannot survive with just accounts and cards
→ Expanding into lending, investment, insurance, and B2B
In other words, from point solutions to bundling
It is also in the conversation.
"Companies that survived the winter have become much stronger than before."
This is unique to finance; because it involves regulation, credit, and cash flow, companies that endure the winter naturally become leaner and more resilient.
1-4. 2025: "The return of spring" = However, the thaw is not over yet
We are currently in "early spring."
Buds are appearing, but snow remains in the background.
And a characteristic of this spring is that another frenzy, AI, is running alongside it.
In other words, the recovery of fintech and the frenzy of AI funding are beginning to intersect.
What is born at this intersection is "AI x Finance," but it will appear not as flashy consumer apps, but as AI that replaces the "operations" of finance.
2. "Access" has been solved. But "quality" has not. The next main battlefield is the substance of finance
2-1. We have digitized. But it has not become excellent finance
The most suggestive part of the conversation is this sentence:
"We have digitized finance, but we haven't necessarily made it 'excellent'.".
The "excellent finance" referred to here is not about the UI on the surface.
It is about whether the logic of finance itself—credit judgment, risk, fraud, underwriting, pricing, and accountability—is in a form that is understandable, convincing, and less likely to cause losses for the user.
Digitization has widened the entrance.
However, there is still significant room for improvement in quality.
That is why the themes leading up to 2026 will be "credit scoring," "fraud prevention," and "agents."
2-2. Embedded finance: Fintech is "everywhere"
Another structural change is embedded finance.
Companies that do not have finance as their core business, such as car manufacturers, agricultural machinery makers, and giant invoicing companies, have started to "embed" financial functions into their own services.
The famous phrase "every company is a fintech company" mentioned here is neither an exaggeration nor hype; it has become a natural part of the experience.
BNPL (Buy Now, Pay Later) will be embedded everywhere, and cards and wallets will be integrated into every experience. As a result, consumers will use financial services without even being aware that they are using a 'banking app'.
3. AI x Fintech in 2026: Three Frontiers
3.1 Redesigning Credit Scores: Evaluating Based on the 'Present' Rather Than the 'Past'
The irrationality of traditional credit scores is felt intuitively, even by those who are not well-versed in finance.
Even when income rises, spending remains the same, and life becomes more stable,
it is not reflected in the score. Even if it is reflected, it is too slow.
This is where Plaid introduces a concept like the 'Lens Score'.
It measures credit 'logically' by looking at income, expenses, and daily cash flow.
This is a renewal of the substance of finance, not just a new app.
If this becomes widespread, what will happen?
Young people, immigrants, and those with weak credit files who have thin credit histories will have a higher chance of being evaluated based on their 'current financial health'
Lenders will be able to grasp risk more quickly and in greater detail
As a result, interest rates and screening processes will move closer to a 'convincing form'
However, this mechanism is also a form of 'surveillance'.
How will extremely sensitive information like income and expenses be handled?
Discussions on financial regulation and privacy will inevitably arise here.
3.2 Agent-Based Finance: A World Where 'You Get a Mortgage by Talking to an AI'
In conversations, this is spoken of quite concretely as a future prediction.
'In two years, talking to an AI app will be the fastest and most efficient way to get a mortgage'.
This is flashy, but the logic is simple.
Mortgages involve a lot of paperwork, complex conditions, and tedious comparisons
Human agents are expensive and have limited processing speed
AI can handle the entire sequence from input to verification, instructions for missing documents, condition comparison, and optimal proposals
However, the biggest barrier is 'trust' rather than 'accuracy'.
As the speaker says, while power users might be able to delegate, it is scary for the average user.
'If I gave this to my mother, she would say, "Where is my money?"'
This distrust is rational.
In finance, a single malfunction can destroy one's life.
Therefore, the key to adoption is not AI acting on its own, but ratherbeing able to explain 'why that decision was made', andthe user being able to stop it.
3.3 The Irony That the Biggest Use Case Is 'Fraud': The Reality of Cat and Mouse
This is the most important and darkest prediction.
"The biggest financial use case for AI is fraudsters using it to commit fraud against financial institutions."
"Financial fraud is increasing at 18-20% per year."
"The cat (the defender) wins in the long run, but right now the mouse (the fraudster) is winning."
This realization can be called the "hidden theme" of 2026.
The more convenient AI becomes, the stronger the attackers become.
Finance, in particular, offers high returns for attackers.
The "pig butchering" type of fraud mentioned in the conversation is symbolic.
It starts with a wrong number or SMS, builds trust, and eventually leads to a money transfer.
In the past, it required human wave tactics ("factories").
However, AI can automate this, and it can be done in multiple languages, 24/7, with infinite patience.
The difficulty here is that fraud is shifting from "technical intrusion" to deceiving human decision-making.
If a user sends money thinking it is "correct," it is difficult for a bank's fraud detection alone to stop it.
In other words, fraud prevention becomes an issue that includes not just model accuracy, but also UX, education, and transaction design.
4. Who will be the winners in 2026: Those who control the "back end of finance" rather than flashy apps
4-1. The TAM has expanded. The "customer" has shifted from consumers to financial institutions
In the latter half of the conversation, the investment theme has clearly changed.
In recent years, the focus has shifted from consumer-facing fintech to software for financial institutions.
The reason is practical.
Customer Acquisition Cost (CAC) has risen
Existing major players are strong, and differentiation is difficult
Meanwhile, there are still mountains of areas within financial institutions where work is tracked using Excel
AI can break into this "unorganized, massive market" for the first time.
This is because AI doesn't just "build screens," it actually does the work.
4-2. Areas where "AI does the work": Compliance, risk, collections, and treasury
The areas listed by the speaker are all mundane, but they are massive and sticky.
Risk/Compliance
Legal
Vendor onboarding
Treasury management
Loan collection/dunning (multilingual, recording, regulatory compliance)
The key point here is that areas where software was previously difficult to sell
may suddenly start selling due to AI.
The reason is that while traditional IT was merely about "efficiency," AI is directly linked to labor substitution (TAM = labor costs).
4-3. The significance of the Board of Directors "understanding" AI
There is another sharp observation.
Changes like cloud migration were difficult for CEOs to understand intuitively.
However, AI is different.
Whether it is a director or a CEO, if they interact with the model, they can experience the "shock" firsthand.
"Board members can type in prompts and intuitively understand the impact."
What does this mean?
Decision-making for corporate adoption will accelerate.
This is because AI is no longer just a "convenient tool" for the front lines, but has become a management issue.
5. The roadmap shown by Plaid: The two pillars of defense (fraud) and evaluation (credit)
5-1. Network-based fraud detection: Cross-sectional data becomes a weapon
The defense strategy Plaid talks about is not just an AI model.
Network-based signals are the core.
Device information
Account data
Behavioral history
And the comparability across "numerous fintechs"
In short, whether "you are anomalous" is measured not just by individual company data, but by the distribution of the entire network.
This is extremely powerful in finance.
Since fraud involves "distributed yet identical attacks," a cross-sectional perspective is effective.
5-2. Redesigning credit: Toward credit that users can accept
The renewal of credit scoring will change the substance of finance.
Moreover, it will ripple out not only to banks but also to BNPL, cards, insurance, and mortgages.
However, at the same time, accountability will be demanded.
"Your score went down because your spending increased."
Even if that seems logical, it can sometimes be felt as pressure by the consumer.
Therefore, in the world of AI credit, not only model performance but also transparency and the possibility of appeal will become competitive advantages.
Conclusion: Fintech in 2026 will enter a phase of "implementation," not "convenience"
The idea that AI will change fintech in 2026 is not about flashy future apps.
Financial "work" will be replaced by AI
Credit evaluation will move toward measuring the "now"
At the same time, fraud will accelerate, making defense the largest growth market.
And, the decision-making process for adoption will accelerate as a management priority.
These four things will move forward in tandem.
Finally, if I were to capture the most realistic atmosphere of this conversation, it would be this metaphor.
"Spring has arrived. The buds are sprouting. But the snow still remains."
AI is a spring breeze, but it is also the cause of a blizzard.
Therefore, the winners of 2026 will not be those who talk about dreams, but those who can step through the mud of the thaw and eliminate the inconveniences of the field one by one.
