AI – Accelerant or a Threat?

For the past several years, disruption has been the dominant anxiety in fintech — who gets displaced, which moats hold, and whether artificial intelligence (AI) renders whole categories of financial software obsolete overnight. At Goodwin’s recent FinTech in Focus panel, a group of investors, operators, and advisers pushed back on that framing. The more interesting question, panelists argued, is not which fintech businesses AI will destroy, but which ones it will turbocharge — and how quickly.

The panel featured Goodwin partner Andrew Davis, Kara Byun of HSBC, Devon Kirk of Portage, Pogos Saiadian of Greyhound, and Jonas Samlin of Synova, and it was moderated by Goodwin partner Arvin Abraham and Gerd Weissenboeck of Zelig. Their conversation covered the AI investment case for fintech, which sectors are attracting capital, the evolving relationship between banks and challengers, and the growing urgency of exit planning in a market that moves faster than it used to.

AI Versus SaaS and Why Fintech Has a Defensible Moat

Investor anxiety about AI tends to cluster around the question of moats. If AI can replicate knowledge work, what stops a well-funded model from displacing a tech software-as-a-service (SaaS) business that took years to build?

The panel’s answer was more reassuring than expected when it comes to fintech. Financial software, it turns out, is among the harder categories of SaaS business to displace. Competitive advantage here is built on deep, highly specific domain expertise — command of particular tax codes, accounting codes, and compliance frameworks. That knowledge is not easily replicated, even with sophisticated AI tools. Regulatory barriers compound the protection. And proprietary datasets, while less unique than they once were as the cost of general data aggregation has fallen, remain difficult to reproduce for fintech businesses.

The important caveat is that subsector positioning matters. Companies sitting in intermediate positions within the value chain — that is, acting as connectors — may find their moats less durable. The firms with the strongest protection are those closest to a regulated activity, with deep customer relationships and switching costs that take years to build.

AI as Accelerant

The panel was more animated when the conversation shifted from defense to offense. AI, the consensus ran, is not necessarily a threat to fintech — it can be a gift to it.

For incumbents, the challenge of legacy technology infrastructure has always been a constraint. Banks carry core systems built up over decades, with layer upon layer of overlapping technology. Fintech challengers can start from a far more efficient base, and AI compounds that structural advantage, enabling product velocity that would have required far more capital a few years ago. Regulatory barriers that once slowed challengers also create an increasingly attractive opening: As regulation evolves, patterns of opportunity emerge, and fintech companies positioned in those spaces move faster than incumbents can respond.

For founders and operators, the implications are equally significant. AI has lowered the barrier to building complex business-to-business (B2B) software companies. A small team with deep domain knowledge can now move at a speed and scale that previously required large engineering organizations. One panelist described a founder effectively running an engineering function with one person and AI tools — a model that would have been inconceivable five years ago. The question the panel left open is whether these productivity gains translate into genuine efficiency or simply enable smaller teams to do more things. The honest answer, for now, is probably both.

It is worth noting that AI did not even make the agenda at the equivalent panel a year ago. That shift — from background consideration to front-and-center strategic question — captures how quickly the landscape is moving.

Where Capital Is Flowing

Panelists identified several subsectors attracting particular attention.

The intersection of stablecoins and payments is generating significant interest. Stablecoins are moving toward the mainstream as payment infrastructure, creating new B2B payment opportunities — particularly for cross-border and institutional use cases. The rise of agentic AI also opens up new payment models that require fresh infrastructure to support them.

The chief financial officer suite remains active, especially in Europe, where financial system fragmentation across jurisdictions creates a clear value proposition for companies solving multi-jurisdictional complexity. Insurance is another area flagged as underserved: Banks have done a considerably better job than insurers in technology adoption, leaving a core system overhaul in insurance largely unfinished.

Across sectors, vertical focus was the investment thesis that resonated most. Own your vertical — even if it is small and specific. Know the patterns, know the customers, and build a business with a clear path to doubling or tripling in size. B2B models were favored for their slower, stickier sales cycles: Once a product is embedded in a client’s workflow, it tends to stay there.

Banks and Challengers: A Complicated Relationship

The panel’s take on the competitive dynamics between incumbent banks and fintech challengers was more nuanced than a simple winner-takes-all framing.

The numbers in favor of challengers are striking. Digital-first providers now account for the majority of banking app traffic, and digital banks hold a meaningful share of UK deposits — up from essentially nothing not many years ago. Apps like Doss in Korea have become top daily-use products in their markets. The pattern is consistent: Consumer attention follows markets, and deposits follow attention.

And yet the relationship between banks and fintechs has become more collaborative than purely adversarial. Large institutions are increasingly comfortable partnering with external companies rather than building everything in house — a significant shift from the dominant model seven years ago. Once a fintech tool is embedded in a large organization, it opens up new touchpoints that are hard to dislodge. The long-run structural question is whether legacy banks can modernize their core technology fast enough. AI is accelerating the urgency of that challenge.

The Exit Conversation, Earlier Than Ever

Exit planning has always been part of the investor conversation. What has changed is how early that conversation is now happening — and how explicitly.

Even at an early stage, founders and investors are thinking about the path to exit — not as a rigid plan but as a way of building in optionality: who the natural buyers are, which relationships need to be cultivated now, and what the product road map looks like through an acquirer’s lens. Panelists were consistent on one point: Being ready is everything. In a market that moves quickly, opportunities arise fast and require fast execution. Running a clean operation, building strategic relationships early, and maintaining a clear-eyed view of whether you are a buyer or a seller in any given consolidation cycle were cited as the foundations of exit readiness.

From a fund perspective, distributions are increasingly front of mind for limited partners— a dynamic that is pushing managers toward earlier and more deliberate liquidity planning. The rise of secondary markets has created new options, and both investors and founders are becoming more comfortable with secondary transfers as a path to partial liquidity ahead of a full exit.

How the Panel Is Using AI Today

The conversation closed with a candid exchange on how participants are actually deploying AI in their own operations. Data room review, document summarization, and first drafts of investment memos are standard use cases, with junior team members often leading adoption. Market mapping, sourcing outreach, and data reconciliation are increasingly automated — with one panelist noting that close to 80% of routine data reconciliation work is now handled by AI.

The tools in use ranged from general-purpose platforms to more specialized applications for market intelligence and note-taking. The consistent refrain was that human review remains non-negotiable. No one is comfortable sending output to a client or counterparty without verification along the way. A centralized productivity suite with appropriate guardrails was seen as essential, particularly given regulatory obligations.

The panel’s final, unresolved question was also its most interesting: Is AI making teams genuinely more productive? The answer, panelists acknowledged, will likely define the next phase of fintech investment as much as any subsector thesis.

See below for a quick recap of our London event:
https://galleries.vidflow.co/c2xlj47v