How mid-sized banks are closing the AI gap

New Delhi, Nov. 12 -- Globally, AI investments in the financial services sector is projected to reach $97 billion by 2027. In India, under the Framework for Responsible and Ethical Enablement of Artificial Intelligence, the Reserve Bank of India (RBI) may allocate a massive public fund of about $575 million towards building shared AI infrastructure and home-grown financial models. Riding on these developments, mid-sized banks are set to harness AI and grow to their potential to turn into mini powerhouses.

Beyond mere AI adoption, the vital question before these banks is how quickly and how deeply they can embed AI-enabled frameworks into their operations. The time for an AI-powered scale-up has come. IDC predicts that by 2027, one in four financial institutions in APEJ will deploy agentic AI to boost efficiency and customer experiences in credit collections in retail lending. And 40 percent of their AI spending will be on new products and services.

Leverage digital technologies India's Digital Public Infrastructure (DPI) is unbundling payments and data to digitize credit. Even as account aggregators and foundational frameworks like Open Credit Enablement Network (OCEN) and Public Tech Platform for Frictionless Credit (PTPFC) take center stage, FinTechs and BigTechs are expanding their presence by leveraging alternative data to bypass traditional collateral requirements.

To stay competitive, mid-sized banks must utilize DPI's foundational digital tools to design tailored, data-driven financial solutions rather than relying solely on traditional lending frameworks. They must adopt alternative data risk-assessment models to evaluate credit where collateralized lending was previously not feasible. Banks must also actively manage the new systemic and cybersecurity risks inherent in digital architectures.

The API banking platform by Federal Bank is a case in point. This Kerala-based mid-sized bank has spent a decade building and integrating robust APIs as an infrastructure effort. The strategic foresight helps the bank offer a range of API services and reap multiple benefits. In fact, the bank's remittance APIs handle over one-sixth of India's total foreign inward remittance.

Build unified data foundations Mid-sized banks face many challenges that prevent AI scale-up, ranging from legacy core systems and thin technology budgets, to the lack of relevant talent and the need to redesign decisions and workflows on AI.

But one of the biggest bottlenecks is data. Many mid-sized banks sit on valuable pieces of information that are trapped in branch systems and spreadsheets, which often don't get integrated into centralized data pools. Such barriers are likely to slow AI experiments down and risk locking mid-sized banks into a shrinking middle.

Data fragmentation is typically a primary reason AI pilots fail to scale. A model trained on a subset of local customer data produces predictions that cannot drive decisions across the full customer relationship.

Banks that are beginning to gain ground often work on a unified data foundation built on cloud-native architecture. They focus on personalized customer engagement powered by AI-led automation. This gives them the competitive edge of being agile with regulatory compliance.

Convert customer proximity into competitive differentiation

While large national banks have the size advantage, mid-sized banks have the advantage of customer proximity, an asset in this age of hyper-personalization. They sit close to their customers through branch relationships, local knowledge, and day-to-day visibility into community needs. So, they have this deep relationship data that AI can work well with.

The use cases are many. Banks can deploy multilingual, voice-enabled assistants that speak local languages and mirror branch-style interactions, making customers feel like they are dealing with a familiar bank officer rather than an inanimate app. Fraud detection and collections can be tuned to community-specific trends like festive spending spikes and local merchant patterns. Alerts can be more precise and interventions more respectful of local practices. Hyper-local product designs are another area where mid-sized banks can leverage the power of AI by using clustering and segmentation models.

Mid-sized banks can achieve rapid scale with AI by selectively deploying high-ROI use cases across multiple areas, while building and maintaining a capital-efficient technology footprint with AI frameworks.

Published by HT Digital Content Services with permission from MINT CIO CIRCLE.