Banks Face Workforce Hurdle in AI Adoption

Most large banks have launched artificial-intelligence pilots aimed at improving customer service, risk management and internal workflows. Yet progress from experimentation to enterprise-wide deployment remains limited. The central challenge is increasingly seen not as access to models or computing power, but whether banks can redesign decision-making, accountability and operating processes so AI — including autonomous AI agents — becomes part of routine business rather than a collection of isolated proofs of concept.

A Capgemini report found that only 10% of financial companies are equipped to deploy AI agents at scale. The finding has sharpened scrutiny of how bank executives frame their AI strategies: focusing narrowly on model accuracy or the return from individual tools can overlook the broader operating-model overhaul required. Scaling will depend on banks aligning governance, data infrastructure, talent and cross-functional workflows with the technology.

Banks have accumulated vast amounts of transaction, customer and risk data over decades, but legacy core systems and data silos limit its use across platforms. Experts argue that banks should not view AI solely as a threat to jobs. Instead, they should use it to improve data connectivity across their infrastructure, turn high-value data into insights and modernize their service architecture.

The latest report urged bank executives to rethink their AI strategies by prioritizing data integration and system connectivity before expanding analytics and service applications. The available information did not identify the experts or institutions behind the views or provide a publication date, investment amount or implementation timetable. The focus therefore remains on the direction of the industry's transformation, with no verifiable quantitative results yet available.

Banks are accelerating AI adoption, rapidly changing the skills required for workflows, risk management and customer service. Traditional training centered on classroom instruction is struggling to keep pace. Most financial institutions still face gaps in AI literacy and practical application, potentially increasing operational, compliance and cybersecurity risks unless they upgrade employees’ capabilities at the same time.

A recent report recommends that banks establish controlled sandboxes where employees can test AI in a secure environment. It also advises cultivating early adopters who can help teams build experience and narrow expertise gaps. The materials provided do not identify specific banks or include investment amounts, survey data or the report’s publication date, so no institution names, amounts or precise timeline can be provided.