Sarah Kim, a partner at Centana Growth Partners, a growth equity fund focused on financial services, shared her perspective on insurtech funding, artificial intelligence and what's ahead with Digital Insurance.
Responses have been lightly edited for clarity.
What are you seeing in the market?
The insurtech marketplace is much more disciplined and honest than it was a few years ago. Capital is still there, but it's no longer chasing growth at any cost. It's rewarding durability, real revenue, and a credible path to profitability.
In that sense, the insurance market is tracking closely with what's happening across the broader technology sector. Valuations are under pressure, exit timelines have stretched, and diligence has returned to levels that frankly should always be the standard. What makes insurtech different is how unforgiving the underlying business is. Insurance isn't a product you can iterate on quickly; it's actuarially complex, heavily regulated, and built on distribution relationships that have taken decades to establish. Technology alone doesn't reduce a combined ratio.
Similarly, AI is front and center in every conversation right now, and the excitement is legitimate. But the pace of change is creating a real strategic problem: Foundation models are improving so rapidly that a capability that felt defensible six months ago can look commoditized today. Every few weeks there's a new model release that reshapes what's possible, and that makes "we built an AI tool" a very thin moat.
The companies with real defensibility are the ones where the value isn't in the model itself, but in what surrounds it. Proprietary claims data is a good example. A carrier or MGA that has embedded years of structured loss data into their underwriting and claims processing workflows has something that may be difficult for a general-purpose LLM to replicate. The model improves, but the data advantage compounds independently.
Add in a more fragile geopolitical backdrop, and the overall tone is cautious and selective. But selective is not the same as pessimistic. The underlying opportunity in insurance technology remains significant.
What type of technology is most promising?
AI is the clear center of gravity, and that's unlikely to change. But the conversation has matured. We're moving past the experimental spend and proof-of-concept phase into substantive deployments. Companies are starting to show what's working in production, and the industry is paying close attention to who's making that transition successfully.
The most compelling use cases are still in
That said, there's still a lot of noise, and the pace of change is rapidly increasing. Investors and executives are hunting for leading indicators of lasting power.
The questions getting asked are: Is the AI embedded in the actual risk decision, or just the workflow around it? Does the product get meaningfully better as more data runs through it? And critically, what happens to this business when the underlying model improves and everyone has access to the same capability? Companies that have good answers to these questions are the ones tending to attract more interest. The ones that don't are finding the conversation much harder.
What kind of M&A deals do you see ahead?
We may be entering into a more active and more pragmatic M&A cycle. We are seeing the "haves and have-nots," where startups that were burning capital without sufficient unit economics are looking for a place to land versus those that add real strategic value and command a premium valuation.
Strategic buyers don't want to be left behind on AI, and in many cases it's faster to buy than build. At the same time, a number of companies are facing a tougher funding environment, which makes M&A a more realistic outcome than raising at a lower valuation. Expect those exits without an AI edge and moat may face headwinds as well with the more sober valuation environment.
We'll likely see a mix of capability-driven acquisitions and opportunistic deals, but in both cases, buyers are far more disciplined on integration and long-term value creation.
How do geopolitical risks affect M&A activity?
They don't stop deals, but they do reshape them.
More broadly, geopolitical uncertainty tends to slow down processes and add friction to cross-border deals. For example, one dynamic worth watching is regulatory divergence. The U.S. and EU are moving in meaningfully different directions on data privacy and AI governance, which raises real questions for insurtech companies operating across both markets. Where do you build? Where does your data live? Which regulatory framework shapes your product architecture? There aren't clean answers yet, but the companies that are thinking about these compliance, governance and regulatory questions early may be better positioned than the ones that assume it will sort itself out.
What hasn't changed is the strategic need for innovation in insurance. If anything, an uncertain environment makes the case for better risk tools more compelling, not less.
Anything else you would like to share?
The hardest part of this moment is that it's genuinely difficult to see around the corner on AI. The pace of change is unlike anything we've seen in this category.
What will matter are proprietary data moats, unique distribution, and talent that can continuously navigate the dynamic market. The teams that win won't just talk about AI. They'll show tangible ROI through better underwriting outcomes, faster claims decisioning, and more efficient operating models.
That gap between narrative and reality is going to define the next phase of insurtech.
