Nikolai Braiden is a seasoned visionary in the financial technology space, having witnessed the early ripples of the blockchain revolution before it became a global tide. As a resident FinTech expert and advisor to high-growth startups, he has spent years at the intersection of traditional banking and disruptive innovation. His perspective is shaped by a deep belief in the transformative power of digital lending and payment systems, but he remains grounded in the practicalities of risk and regulation. In this conversation, we explore how the industry is moving beyond the initial excitement of generative AI to embrace a new era of autonomous agents that don’t just talk but actually perform the heavy lifting of financial engineering.
The dialogue delves into the massive shift from generative tools to agentic AI systems, comparing this evolution to the invention of the automobile. We explore the staggering productivity gains being reported by industry leaders and the critical necessity of rigorous data governance to prevent the “unacceptable” risk of hallucinations in banking. Braiden highlights why the ceiling for AI in finance remains exceptionally high, emphasizing that the most successful implementations are those that pair cutting-edge automation with human judgment and a steadfast commitment to trust.
The financial world has moved rapidly from the initial shock of ChatGPT’s launch in 2022 to a more structured adoption of AI. How do you interpret the industry’s pivot from basic generative tools toward these more autonomous agentic systems?
The transition we are witnessing is nothing short of a fundamental re-engineering of the financial plumbing. When ChatGPT arrived in late 2022, the industry was captivated by its ability to generate text, but the real magic is happening now as we move toward agentic solutions that can actually execute complex workflows. According to the 2026 Global AI in Financial Services Report from the University of Cambridge, a significant 81% of respondents expect these agents to be mainstream by 2030, which tells you just how much momentum is building behind the scenes. It is telling that 52% of institutions have already started integrating this technology into their stacks, moving away from simple “chatbots” to systems that can hypothesize and verify. For me, the sensory experience of this shift is like moving from a noisy room of people talking to a silent, highly efficient laboratory where work is getting done without constant supervision. It is about moving from a tool that helps you write a letter to an agent that can manage a multi-step dispute process from start to finish.
There is a striking comparison being made between the rise of agentic AI and Henry Ford’s development of the automobile over faster horses. How does that analogy resonate with the way you advise startups to rethink their core processes?
The “faster horse” analogy is perfect because it highlights the difference between incremental improvement and total transformation. In the past, we were just trying to make existing banking processes move a bit quicker, but now we have the opportunity to throw out the old blueprints entirely and build something new. Last year, many firms were excited about a 2x to 3x jump in productivity, which is respectable, but this year the conversation has shifted toward a potential 100x increase in output. When you see numbers like that, you realize you aren’t just making the horse run faster; you’ve built an engine that changes the scale of what is possible. For a startup, this means you can build products that were previously too expensive or complex to manage, like personalized wealth management for the average person. It requires a certain amount of bravery to abandon the “horse” of legacy systems, but the institutions that do so are the ones that will find themselves leading the market in three to five years.
When we look at the mechanics of these agents—reading documents, researching databases, and verifying hypotheses—how does this change the day-to-day reality for the thousands of employees working in global banks?
The impact on the workforce is profound, as seen with institutions like Santander extending AI capabilities to 185,000 employees worldwide. We are moving toward a reality where the AI does the “grunt work” of data gathering and preliminary analysis, allowing the human to act as the final arbiter and ethical guide. Imagine an agent that doesn’t just present you with a pile of data but actually investigates the database, creates a hypothesis about a fraud pattern, verifies it against historical records, and then presents a fact-checked answer. This removes the mental fatigue of manual data entry and the “hallucination” risks that come with human error or basic generative tools. It feels like having a highly skilled research assistant who never sleeps and who provides a level of responsiveness for customers that was previously unimaginable. The goal isn’t to replace the human touch but to give our teams better insights so they can be proactive rather than reactive in their service.
Trust is the bedrock of banking, and many experts have called GenAI hallucinations “completely unacceptable” in this sector. How are institutions practically addressing the need for responsible AI and data governance?
In banking, there is no room for “creative” errors; a hallucination isn’t just a glitch, it’s a potential regulatory nightmare and a breach of trust. This is why we are seeing an intense focus on correcting data, metadata, and the descriptions of items within databases to ensure the AI has a solid foundation to work from. Leaders in the space are being very intentional about how these tools are secured and aligned with their core values, especially in credit unions where the human connection is the primary differentiator. If the data architecture is flawed, the AI will build a house of cards, so the real work right now is in the “unsexy” side of technology—data cleaning and governance. We have to pair these powerful tools with strong human judgment to ensure that every decision made by an agent is transparent and accountable. It’s about building a digital environment where the customer feels secure knowing that while an AI might be processing their request, the bank’s integrity and human oversight remain the guiding lights.
The Banking Tech Awards highlighted that some of the most impactful innovations are coming from specialist companies rather than the massive global card networks. Why are these smaller players often better positioned to lead in AI?
There is a common assumption that only the “Goliaths” of the industry have the resources to innovate, but specialized companies often have much deeper domain expertise and access to highly relevant, niche data. For example, a company like Chargebacks911 can use their Unified Dispute Management System to surface operational insights in real-time because they are laser-focused on one specific, complex problem. These specialists aren’t bogged down by the same level of legacy bureaucracy that slows down a massive bank, allowing them to act on data more consistently and effectively. They are proving that the most successful AI isn’t always about having the largest model, but about applying existing capabilities to the right data to solve real-world friction. When these smaller firms win awards for data insights or intelligent workflows, it serves as a powerful validation that expertise and agility are just as valuable as scale in the AI race.
As we look toward the future, many believe we haven’t even come close to the “ceiling” for AI in financial services. What does a “fully intelligent and adaptive” banking environment actually look like to you?
The ceiling in banking is incredibly high because the industry is built on complex variables like risk, regulation, and personalized relationships. A truly adaptive environment isn’t one that is just fully automated and cold; it’s one that is more responsive and relationship-driven because it anticipates needs before the customer even articulates them. We are moving toward systems that reduce friction at every turn, using AI to identify financial needs earlier and deliver personalized service at a scale that was once impossible. This environment will be one where AI expand across every operation—from risk management to software delivery—while maintaining a high bar for accountability. The winners in the long term will be those who use technology to strengthen the human connection, creating a banking experience that feels both futuristic and deeply personal. It’s about creating a world where the technology fades into the background, leaving only a seamless, intelligent financial life for the user.
What is your forecast for agentic AI in the financial sector?
I forecast that by the time we reach the end of this decade, the very definition of a “bank” will have shifted from a place where you store money to an autonomous financial partner driven by agentic AI. We will see the 81% mainstream adoption rate predicted for 2030 manifest as a total disappearance of manual back-office processing, with agents handling everything from complex credit disputes to real-time risk assessment. The productivity gains we see today, moving toward that 100x mark, will allow institutions to offer hyper-personalized financial advice to every single customer, regardless of their net worth. However, this progress will be strictly gated by regulation; the institutions that fail to master data governance and “fact-checked” AI today will likely be left behind or consumed by those who did. Ultimately, the future of banking won’t be a cold, robotic interface, but a highly intelligent, proactive, and deeply trusted relationship where the AI acts as the engine and human values act as the steering wheel.