As artificial intelligence moves from a promising experiment to essential infrastructure, investors are becoming much more discerning about what truly counts as an AI-native company. By 2026, the gap between startups that simply plug in AI tools and those built around AI from day one is widening fast and nowhere is that more obvious than in heavily regulated industries like finance.
An investor with a front-row seat
Dmitry Volkov has had a front-row seat to this shift. A serial entrepreneur and investor, he was an early backer of companies such as OpenAI, Revolut and Patreon. Over the years, Volkov has invested more than $500m across more than 20 ventures. Now, he’s placing his bets on what he believes is the next phase of fintech: banks built with AI at their core.
Building a bank around AI
Through his latest venture, Molit.ai, Volkov is supporting the creation of a European bank designed around artificial intelligence, not as a feature, but as its underlying operating system. For him, the project reflects a broader change in how investors think about AI businesses — and what they’re now willing to fund.
Why data alone is no longer enough
One of the biggest shifts, Volkov says, is how data is judged. A few years ago, having vast amounts of data was often seen as a defensible advantage. Today, that’s no longer enough. “What really matters,” he explains, “is whether the data is proprietary, legally protected, and generated through real product use.” Aggregated or scraped datasets, once common in early AI startups, are increasingly viewed as weak foundations.
The end of patient capital
Investor patience has also worn thinner when it comes to monetisation. Startups are now expected to show early clarity on how they’ll make money, rather than relying on long-term stories. “Founders have to be very precise about the problem they’re solving,” Volkov says. Vague or overly broad AI ambitions, he adds, are much harder to justify in a crowded market.
Product execution takes centre stage
At the same time, building AI models has become easier, intensifying competition. As a result, differentiation is shifting away from the technology itself and toward how it’s applied. The strongest teams Volkov sees today are deeply product-focused, using AI to solve specific, real-world problems rather than chasing general-purpose capabilities.
Why banking needs a reset
This thinking explains his belief that banking is due for a deeper reset. While neobanks have refreshed interfaces and distribution, Volkov argues that most financial institutions are still weighed down by systems designed decades ago. “You can’t fix that by bolting AI on top,” he says. “Those limitations run through everything.”
Banking as an ongoing partnership
Molit.ai is built around a different premise: treat the bank itself as a technology-native system, with intelligence woven directly into its foundations. That approach reframes banking from a list of features into an ongoing relationship, where an AI-driven system understands users’ habits, preferences and financial needs.
Rethinking customer loyalty
Volkov believes this could transform customer loyalty. “Feature count is overrated,” he says. “What matters is how services are delivered.” When banking becomes easier, more timely and less intrusive, it fits more naturally into everyday life — and stops feeling like a chore.
Regulation, not rejection
Regulation, of course, remains a major hurdle. But Volkov rejects the idea that AI-first models weaken trust or oversight. When designed properly, he argues, these systems can actually strengthen compliance by improving consistency, transparency and auditability.
AI with humans in the loop
“Being AI-first doesn’t mean taking humans out of the loop,” he says. “It means helping them make better decisions.”
Support as part of the product
That philosophy also guides Molit.ai’s approach to customer support. Instead of using AI to deflect queries, the system is designed to help proactively, drawing on context and history to resolve issues quickly. “Too many banks see support as a cost centre,” Volkov says. “We see it as central to the product.”
Advice for the next wave of founders
For founders building AI-first startups today, his advice is straightforward: treat AI as infrastructure, not decoration. “No matter how advanced the technology,” he says, “long-term success still comes down to understanding who pays, why they pay, and how that scales.”
A defining divide for fintech
As investor expectations rise and regulatory scrutiny tightens, the divide between truly AI-native institutions and those retrofitting intelligence onto old systems is likely to shape the next chapter of fintech.
