If you look under the hood of European wealth and private banking today, the contrast is pretty brutal. Client portals and mobile apps have had a facelift; the back end for IFAs, family offices and private banks often still lives in spreadsheets, legacy desktop software and manual reconciliations. Relationship managers and ops teams improvise around the gaps and everyone pretends the core is “good enough” because ripping it out feels impossible.

Meanwhile, the economics are moving the other way. Europe’s wealth-management industry sits on roughly €37 trillion of assets, but revenue margins keep tightening and regulatory overhead keeps rising. You cannot run that scale of assets on 1990s tooling forever. The global WealthTech market is projected to more than triple this decade, and most of the serious growth is on the B2B side: infrastructure for banks, wealth managers and family offices rather than direct-to-consumer apps.

We’re already seeing what that looks like. Data-layer platforms quietly pipe positions and transactions across hundreds of thousands of portfolios for dozens of institutions and become de facto operating systems without ever touching the end client. In capital markets, automated surveillance, risk and reporting layers are starting to sit on top of very old core systems, reading everything and flagging what matters in real time. Fintech funding is consolidating in the same direction: fewer deals, larger rounds, and a bigger share of capital flowing to back-office and market infra rather than yet another neo-broker or savings app.

Insurance has long trailed banking in technology rollout, but that gap is exactly where the real opportunity now sits. The operational core is largely untouched and it’s now finally opening. Buyers want automation where the economics live: claims, underwriting, servicing and compliance. That’s why the teams breaking through are wiring AI directly into the data and workflow layers of insurers and financial institutions.

In parallel, fintech credit is going through its own evolution. For years, lending startups were dismissed as too risky or capital intensive. The difference now is the data: live feeds from banking, payments, portfolios and ERP systems make it possible to price risk dynamically in real time, rather than off stale PDFs and static scorecards. With that, entire lending niches start looking more attractive because they can finally be underwritten with real precision.

Insurance, in particular, is an almost perfect AI playground. It runs on messy, unstructured data: handwritten notes, PDFs, medical reports, photos, call transcripts. Five to seven years ago, OCR was the innovation ceiling. Today, you can feed the entire file to a model and get a proposed decision with an auditable trail. Early deployments are already cutting handling times from weeks to days and taking meaningful double-digit percentages out of processing costs. On the credit side, we’re seeing the first generation of players use real-time cashflow, portfolio or treasury data to underwrite with far more precision, rather than throwing capital at growth and hoping defaults behave.

My expectation for 2026 isn’t a new insurtech hype cycle. It’s a quieter wave of infrastructure-like businesses selling into insurers, banks and CFOs, fixing specific P&L problems, and using AI because it actually changes the economics, not because it looks good in a pitch deck.