Contrary to expectations, the big banks are not the only runaway leaders when it comes to unlocking AI value. The latest Infosys Bank Tech Index: Retail and regional banks, finds that retail and regional banks are holding their own, achieving AI business value at rates comparable to their larger peers. Retail and regional banking respondents said that approximately 59% of their deployed AI initiatives generate measurable business value, up eight to ten percent points from the year before, and in line with the overall banking sample.
How are the retail and regional banks achieving this despite fewer resources and a smaller scale of operations? These banks have understood an important fact: AI success closely tracks core modernisation spend. Thus, 90% of them are committing at least 10% of their technology budgets to drive core system modernisation; this includes 68% who are spending between 20% and 39% of their total technology budgets on modernising or replacing core banking systems.
Modernised core+AI: Essential pairing
For banks running on legacy infrastructure, core modernisation typically used to be their way out of mounting technical debt and operational friction. Today, it is also the way to scaling AI across the organisation. A mutually reinforcing relationship is setting in: even as banks increasingly upgrade their core systems to support AI capabilities, other AI tools are helping them modernise with greater efficiency.
Traditional core systems are tightly integrated monoliths, built for batch processing rather than the continuous, real-time data streaming AI applications need. Modern platforms, by contrast, are composable and microservices-based: components can be modified individually, often without disrupting the entire system, while zero downtime updates and continuous integration and deployment ensure uninterrupted real-time processing at scale. The best platforms are cloud-native, cloud-neutral and multi-cloud, essential traits for running AI technologies.
For example, in a legacy core, each product or service line such as loans, deposits or payments is served by a separate data source, often leading to duplication, inconsistency and other problems that greatly hinder AI adoption. Modern platforms house their data in a central repository that is managed independent of applications and processes. Acting as a single source of truth for the enterprise, the repository facilitates the adoption of analytics and AI applications across banking operations. Advanced core banking solutions also offer enterprise and open-source options, along with enterprise-grade predictive, generative and agentic AI platforms for scaling AI journeys. A critical reason for switching to a modern core platform is its multi-layered security architecture, built in accordance with Zero Trust principles. This is a growing necessity as agentic AI systems require broader access across the data estate, expanding the surface that banks must monitor and protect.
Around the world, banks are increasingly refreshing their core systems to gain competitive advantage. Royal Bank of Canada offers an instructive industry example. RBC Clear, its cash management platform for corporate clients, was built cloud-native and microservices-based from the ground up. All components of the platform are cloud-native and microservices based, allowing RBC to benefit from the cloud’s elasticity, availability and automation, along with faster delivery and feedback cycles. That architecture did more than modernise the plumbing, it enabled new AI-driven capabilities layered on top. A payment orchestration engine sitting atop the bank’s payment rails provides a pre-screening capability that runs every payment through sanction, fraud, AML and credit-limit checks before it is sent to the relevant rail, directly addressing a pain point RBC had heard repeatedly from corporate treasury teams: time spent chasing down payment status. The cash management business became, in effect, a proving ground for capabilities that RBC could extend elsewhere across the enterprise.
AI returns the favour
Core modernisation is a prerequisite for AI adoption. But now, AI is increasingly becoming a fixture in core systems development. Progressive organisations are leveraging AI not only for writing and testing code, but also for evaluating legacy systems, extracting business rules and automating entire processes. This is helping to prevent costs and timelines from overshooting their targets, often a major concern in modernisation programs.
Vendors are building AI in from the start
As core banking and AI systems increasingly become mutually dependent, core platform providers are integrating data and AI within their foundations to help banks scale AI with confidence. Leading platforms support the entire spectrum of enterprise AI, including data readiness, predictive insights, generative intelligence and autonomous agents, empowering global and regional banks, and also those in-between, to adopt and scale AI.
The trend line from this year’s index is worth sitting with: retail and regional banks closed most of the gap with larger peers in a single year, and they did it by treating core modernisation as a prerequisite rather than an afterthought. As that playbook spreads, scale may matter less to AI success than the willingness to rebuild the foundation underneath it.
Ajay Bhandari, Senior Vice President of Financial Services at Infosys
