Hi! Interesting read. It's absolutely spot on considering the fact that critically regulated industries such as healthcare, financial services and legal, trust continues to be the major hurdle for adoption. This is important because guardrails are required with respect to bias, transparency and fairness. Moat does not mean technological defensibility but rather the ability to understand your customer's context and build your product with a focus of winning the customer's trust.
I write a blog in substack titled "The LegalTech Thesis" wherein I analyze LegalTech startups and identify opportunities for investment. Would love to get your thoughts on my post!
this is the one that keeps getting validated. across the VC newsletters we track, the funds backing AI companies with real traction keep saying the same thing. thank you for the article!
Victoria, this is an excellent piece — and the timing is particularly relevant given where we are in the AI adoption cycle within financial services.
Your distinction between building "for" an industry versus building "in" an industry is one we encounter constantly in our advisory work with banks, asset managers, and fintechs. The firms that understand the muscle memory of how a treasury operation or compliance function actually works have an enormous advantage over those that treat financial services as just another vertical to bolt AI onto.
I'd push the thesis one step further, though. Trust isn't just a moat for AI-native startups — it's becoming the decisive competitive variable for incumbent financial institutions as they navigate the convergence of AI, tokenized assets, and programmable settlement infrastructure.
In our research on the "Convergence Economy," we've found that the simultaneous compression of intelligence costs (AI) and coordination costs (stablecoins, smart contracts, tokenized rails) is expanding the competitive surface area dramatically. The question shifts from "who can build the technology?" to "who do counterparties, regulators, and clients trust to operate it responsibly?" This is why we've argued that compliance is a moat, not a barrier — the GENIUS Act's requirements are substantial, but for institutions with existing regulatory infrastructure, they represent a defensible advantage that unregulated entrants can't replicate.
Where this gets most interesting is at the intersection of agentic AI and tokenization. When autonomous AI agents begin executing financial workflows on programmable rails, the trust question becomes existential, not just competitive. Your point about restraint as a trust signal maps directly: knowing when your AI agent should route to a human rather than hallucinate an answer is the institutional equivalent of knowing when to say no.
We've published research on these dynamics — multi-agent AI design models for financial institutions, the convergence of AI and digital asset infrastructure, and the strategic imperatives that flow from this shift — at nextfiadvisors.com/intelligence. Happy to share if useful.
Great piece. This is exactly the kind of thinking needed as these industries navigate what is fundamentally a trust-intensive transition.
Hi! Interesting read. It's absolutely spot on considering the fact that critically regulated industries such as healthcare, financial services and legal, trust continues to be the major hurdle for adoption. This is important because guardrails are required with respect to bias, transparency and fairness. Moat does not mean technological defensibility but rather the ability to understand your customer's context and build your product with a focus of winning the customer's trust.
I write a blog in substack titled "The LegalTech Thesis" wherein I analyze LegalTech startups and identify opportunities for investment. Would love to get your thoughts on my post!
https://harshithviswanath.substack.com/p/three-legaltech-whitespace-plays?r=4y4gfu
this is the one that keeps getting validated. across the VC newsletters we track, the funds backing AI companies with real traction keep saying the same thing. thank you for the article!
Victoria, this is an excellent piece — and the timing is particularly relevant given where we are in the AI adoption cycle within financial services.
Your distinction between building "for" an industry versus building "in" an industry is one we encounter constantly in our advisory work with banks, asset managers, and fintechs. The firms that understand the muscle memory of how a treasury operation or compliance function actually works have an enormous advantage over those that treat financial services as just another vertical to bolt AI onto.
I'd push the thesis one step further, though. Trust isn't just a moat for AI-native startups — it's becoming the decisive competitive variable for incumbent financial institutions as they navigate the convergence of AI, tokenized assets, and programmable settlement infrastructure.
In our research on the "Convergence Economy," we've found that the simultaneous compression of intelligence costs (AI) and coordination costs (stablecoins, smart contracts, tokenized rails) is expanding the competitive surface area dramatically. The question shifts from "who can build the technology?" to "who do counterparties, regulators, and clients trust to operate it responsibly?" This is why we've argued that compliance is a moat, not a barrier — the GENIUS Act's requirements are substantial, but for institutions with existing regulatory infrastructure, they represent a defensible advantage that unregulated entrants can't replicate.
Where this gets most interesting is at the intersection of agentic AI and tokenization. When autonomous AI agents begin executing financial workflows on programmable rails, the trust question becomes existential, not just competitive. Your point about restraint as a trust signal maps directly: knowing when your AI agent should route to a human rather than hallucinate an answer is the institutional equivalent of knowing when to say no.
We've published research on these dynamics — multi-agent AI design models for financial institutions, the convergence of AI and digital asset infrastructure, and the strategic imperatives that flow from this shift — at nextfiadvisors.com/intelligence. Happy to share if useful.
Great piece. This is exactly the kind of thinking needed as these industries navigate what is fundamentally a trust-intensive transition.
— Barry Eisenberg, NextFi Advisors