Europe’s financial sector is experiencing an “internet 1996” moment with artificial intelligence (AI): Everyone can see the direction of travel, yet the gap between pioneers and followers is already visible. If the followers don’t step up their efforts soon, they could be left behind permanently.
The latest Oliver Wyman Forum survey shows that 17% of NYSE-listed companies — and 16% of European respondents — have already achieved revenue or cost improvements of more than 10% from AI initiatives, and these leaders are five times likelier than peers to say projects exceeded return on investment expectations (25% versus 5%).
The opportunity is vast because the areas where incumbents dominate — trust, strong balance sheets, and proprietary data — are exactly what AI can amplify. Customers still entrust their salaries and savings to established firms, reducing the firms’ funding costs, because of their robust regulatory oversight and risk management. That trust yields a torrent of granular data which, if activated, could become the raw material for personalized offers before open-banking interfaces fully level the playing field.
In practice, however, many banks, insurers and asset managers continue to treat AI as a series of proofs-of-concept rather than as a strategic operating system as they await perfect regulation, flawless AI models or both. The risk is that market share and cost advantage will drift to competitors. Fintech and big tech have already captured material share, especially in retail brokerage and payments, where some fintechs are seeing year-over-year growth between 25% to 34%.
Embracing customer-centric design for business success
For established firms, cultural change sits at the heart of building a credible AI roadmap. They face the organizational hurdle of remaining wired around product silos while AI rewards customer-centric design and fluid data usage across the organization. That obstacle must be cleared.
Equally important is adopting a “marginal-cost-to-zero” mindset. That will turn efficiency into growth and will ultimately also benefit the broader European growth agenda.
Lending is the most vivid illustration: When marginal cost approaches zero, lenders can viably serve thinner-margin segments such as startups, gig-economy workers, and micro-subject matter experts (SME), which legacy processes have priced out of the market. At one European bank we have worked with, an AI engine that pre-underwrites small business loans by combining transactional flows, point-of-sale receipts, and sector benchmarks has cut unit underwriting costs by roughly 70%, shrunk decision times from days to minutes, and expanded the addressable market enough to lift SME lending revenue by about 50%.
Numerous execution enablers separate AI leaders from laggards. First, leaders treat data as a product with clear ownership, service-level agreements and roadmaps; without that discipline, even state-of-the-art models disappoint. Second, their vendor strategy balances speed against dependence: Black-box demonstrations, “100% accuracy” claims, or unclear model lineage are warning signs. Third, reliability economics matter, because every additional “nine” of availability —from 90% to 99,999% — requires exponentially greater investment, and boards must decide where full autonomy is genuinely worth the spend.
Reskilling the AI workforce is key to scaling enterprise adoption
AI leaders understand that their workforce models must evolve, too. CEOs overwhelmingly expect AI to raise productivity rather than cut headcount, yet staff reskilling for prompt engineering, model governance, and human-in-the-loop supervision is still behind the curve. Employees at AI-leading firms are three-and-a-half times likelier to use AI daily and twice as likely to say it has changed collaboration, creating a cultural flywheel that accelerates adoption.
Leaders are already wiring these elements into execution. In all, 71% of AI-leading CEOs cite technology as the prime driver of long-term competitiveness, compared with 45% of non-leaders, and almost a third of leaders already generate more than 20% of revenue from AI-enabled products. They involve their boards more actively, especially in engaging regulators and investors, and favor organic capability building over acquisitions as the path to shareholder value. They also tie success to metrics that cut through hype: AI revenue share, marginal cost per transaction and employee engagement.
Why banks must scale AI beyond pilots to stay competitive
Companies still in pilot mode or taking a wait-and-see approach need to accelerate their efforts. They should articulate a North Star ambition anchored in the business model of the future, not a patchwork of AI widgets bolted onto legacy architecture. They should engage supervisors early so that governance becomes a competitive advantage rather than a retrofit tax. They must track the metrics that leaders already watch, and they must mobilize boards to own the external narrative.
Generative AI won’t make banks obsolete; it will make some banks extraordinary. Those that fuse hard-earned trust, proprietary data and disciplined execution into AI-native operating models will extend their advantage rapidly, while those that hesitate will lose ground by the quarter. For Europe’s financial champions, the moment to choose between pilot and profit is now.
Read the original piece here.