Why Regulation Is Your Greatest Moat: A FinTech Founder's Playbook for AI & Innovation

Originally published in Forbes.

Many of today's FinTech founders share a similar origin story. We started our careers on Wall Street, in the heart of "traditional finance". We learned the hard and soft skills of a demanding industry, but we also saw its deep inefficiencies. We then left to build something new, driven by a common mission: to apply technology to inefficient markets and democratize access to asset classes, from hedge funds to real estate to private equity, that were historically cornered off.

This "TradFi-to-FinTech" journey provides a unique lens: an insurgent's desire to innovate, paired with an incumbent's deep-seated understanding of risk.

Today, that dual perspective is more critical than ever. We are facing the most powerful, and the most hyped, technological shift since the dawn of the internet: Generative AI. For FinTech founders, this technology presents a dangerous paradox. It is simultaneously a massive accelerant and a potential business-ending liability.

Navigating this new era requires balancing the hype with a high level of responsibility. For founders looking to build an enduring company in this space, here is how I believe we must approach the new frontiers of AI in regulated markets.

The AI Paradox: Internal Gains vs. External Dangers

Internally, AI is already an unambiguous win, offering massive efficiency gains.

• Product teams are using AI-powered design tools to go from idea to prototype in record time.
• Customer support teams are using AI trained on internal knowledge bases to provide instant, accurate answers, deflecting simpler queries so humans can focus on complex issues.
• Engineers are using co-pilots to increase velocity, automating pattern-based code to focus on scalable architecture.

These internal tools allow fintech companies to iterate and ship valuable tools to clients faster. However, the moment this technology directly faces the end client, the stakes change.

The core challenge is that today's large language models (LLMs) are non-deterministic. They can, and do, hallucinate. In a high-stakes, regulated industry like finance, this is an unacceptable risk. You cannot have an AI tool give a client a hallucinated answer about an investment they are about to make with their hard-earned money.

The 100% Problem: Why FinTech Needs Airplanes, Not Cars

In the tech world, we often talk about 99.5% uptime as a goal. But in finance, that's not good enough.

While some of my peers that started their careers at big banks can say, "JPMorgan will always be there", I can't. I worked at Lehman Brothers. I learned firsthand that you cannot be "mostly" right about risk. This is why I advocate for a controlled method of rolling out AI: use it aggressively for internal staff productivity, but always gate a human between the AI's result and the end client.

Think of it this way: some technology is like a car. You can accept a certain level of risk for a 99% safe trip. But in FinTech, when you are deploying people's capital and managing their nest eggs, you are not building a car. You are building an airplane. You don't need 99% of airplanes to land safely; you need 100%. Incidentally, I previously worked at Lockheed Martin in a group responsible for keeping planes in the air, where I learned of the “9’s rule”. The software tracking planes has to strive for 99.9999999% reliability.

For a financial platform, a single mistake that leads to a financial loss for a client is a complete non-starter. The only responsible solution, until the technology is 100% reliable, is to deploy it in a narrow or gated way. With this approach, you have to accept that it may create dead ends for users, rather than giving a potentially wrong, hallucinated answer. While it may be a degradation in user experience in the short term, it is the non-negotiable price of trust and accuracy.

Regulation Is a Moat, Not a Nuisance

This cautious approach leads to the second pillar of building an enduring FinTech: embracing regulation. Many founders shy away from regulated industries, seeing them as a barrier to innovation. This is a mistake. In my experience, these constraints are actually a powerful strategic advantage.

First, constraints breed creativity. Working within guardrails forces you and your team to be more creative and more focused, leading to a better, more resilient product.

Second, regulation is a competitive moat. Solving hard problems in a compliant way is difficult, which means fewer competitors are willing to do the work. It weeds out those who aren't serious and builds immense trust with clients who see you are doing things the right way. It also builds trust with incumbents who have established distribution channels you may want to utilize down the line.

Finally, you cannot reach the mass market without regulation. Uncertainty means high cost of capital, and regulation reduces uncertainty by offering rules of the road, however imperfect. The "early adopters" who love startups are, by definition, risk-takers. But they are a tiny fraction of the market. The majority of the population (the mass market you need to build a truly scaled business), is thankful for the guardrails that regulation provides. Therefore, if you treat regulation as a nuisance, you are signaling to 80% of your potential market that you don't share their values, or align with their risk tolerance.

The Final Lesson: Accessibility Wins the Race

As founders, we are constantly looking at the next horizon. Right now, the two biggest are AI and blockchain. But their adoption curves have been radically different. AI, specifically LLMs, has had a much quicker adoption cycle. Why? Accessibility.

Companies like OpenAI and Google made their technology magical by meeting users where they are. This allowed ChatGPT to go from an idea to a market mainstay in three years. They gave us a simple chat box and a wow factor that sparked the imagination of mass consumers. Blockchain, while powerful, remains largely inaccessible, requiring a high degree of technical understanding to even use, let alone trust. Bitcoin is going mainstream but it's been building market credibility for over 17 years since it entered the market in 2008.

The lesson is simple: meet your clients where they are. Change is hard, and underestimating the friction it causes is a classic startup mistake. Whether you're a technical or non-technical user, the products that win are the ones that integrate into your life seamlessly, not the ones that demand you change it.

Building an enduring FinTech company is a journey down a harder path. It requires a commitment to learn, a willingness to be technical, and, above all, the discipline to prioritize trust over hype. However, the challenges are worth it. The gratification of driving access to new markets and providing the opportunities for growth and financial freedom that comes with it, has no ceiling.