RBI Deputy Governor Pushes for Stronger AI Governance in Financial Institutions

“Investment in digital infrastructure and a data-driven approach while subverting risks of heavy dependence on third-party technology providers allows financial institutions to remain competitive'” Reserve Bank of India (RBI) Deputy Governor, M Rajeshwar Rao remarked recently.Rao was giving the inaugural address at the IIM Kozhikode and National Stock Exchange joint Second Annual Conference on Macroeconomics, Banking, and Finance and spoke about the role of digital transformation in finance and the use of artificial intelligence (AI) and machine learning (ML), among other topics. Here are some interesting points Rao discussed:Digital transformationRao centred his speech around “creative disruptions” or the dismantling of old systems and their replacement by new ones, specifically in the financial sector.

The ‘India Stack’ program comprising innovations like the Unified Payment Interface (UPI), the Account Aggregator (AA) framework, and the Unified Lending Interface (ULI), supplementing the traditional banking system and expanding financial reach, promoted such disruption, he claimed.Rao also noted that digital lending has helped integrate underserved segments into the formal financial system. His comments echo RBI’s broader ULI push, as evident in the recent meeting between the RBI Governor and MDs and CEOs of select non-banking financial institutions (NBFCs), wherein the latter were urged to adopt the ULI framework.



Previously, RBI had explained that ULI would reduce credit appraisal time for smaller and rural borrowers alongside digitising access to the customers’ financial and non-financial data.spurring financial inclusionBesides ULI, UPI also enables greater financial inclusion, especially for the informal sector, initially dealing in cash, Rao explained. UPI’s ubiquity creates financial footprints for members of the informal sector, which can be used by financial service providers (FSPs) in the account aggregator framework.

These financial footprints also help extend credit to new-to-credit individuals, gig workers, and small businesses without formal credit histories.Recently, the central bank extended credit line facilities (pre-approved borrowing limits) via UPI for small finance banks, aiming to improve access to short-term credit for small businesses. This adds to RBI’s provision of constant support to the UPI framework through acts like levying zero merchant discount rates (MDR) and even having spent Rs 3,600 crore on its promotion (and that of RuPay) between 2021-24.

AI/ML driving the next phasesRao highlighted the growing discussions on AI adoption among financial institutions, including commercial banks and regulators. Further, he outlined three key areas where AI adoption could aid financial services- risk assessment & credit scoring, enhancing customer experience, and fraud detection. AI-driven models can analyse copious amounts of data like transaction patterns, utility bill payments, and e-commerce behaviour to establish a borrower’s creditworthiness and also predict stress among existing borrowers, thereby formulating remedial measures.

Additionally, financial institutions can leverage AI to offer customers hyper-personalised financial products, bettering their experience. Finally, AI-based techniques enabling fraud detection are critical for real-time payments and digital transactions, where cyber threats and frauds loom, Rao opined. Previously, RBI’s subsidiary unit Reserve Bank Innovation Hub (RBIH) launched MuleHunterAI, a tool to detect and flag mule accounts enabling financial fraud.

The move came after RBI noted discrepancies like high false positives and longer turnaround times in the older static-rule-based system. Where does AI lack?lack of explainabilityRao stated that the challenges posed by AI like algorithmic bias, data privacy, and security stem from a “lack of explainability”. Elaborating on this, he noted that since AI systems function like black boxes, producing outputs that are difficult to interpret for their developers, this system falls short in the financial sector, driven by trust, accountability, and regulatory compliance.

Barring this explainability, human intervention acts as “rubber stamping” instead of responsible oversight, he concluded. Moving forward, while AI systems undergoing ‘dynamic adaption’ are beneficial, they can increase occurrences of data drifts (a situation where the data inputs a system is trained on differs from the data inputs it is applied to) and concept drifts (a situation where data evolution invalidates the model). Rao claimed that such models risk providing incorrect financial decisions and instability and must be countered with regular human oversight and explainability.

Further, he noted that the shortage of professionals to oversee AI models exacerbates the challenges associated with the lack of explainability.over-reliance on aiMoreover, “automation complacency” or excessive reliance on technology despite certain situations requiring careful judgement is another risk of AI-based decision models, Rao added. Steps to take going forwardWhile acknowledging the benefits of digital transformation in the financial sector, Rao explained that concerns around excessive borrowing have been on the rise.

To counter this, he called upon financial entities to ensure increased financial literacy and consumer awareness concerning the risks of leveraged products and “speculative investing”. Concerning the integration of AI, cloud computing, and API-driven finance, Rao explained that financial institutions must install robust governance and risk management frameworks. Going forward, Rao remarked that entities must prioritise central risk awareness, ethical AI usage, and customer-centric innovation, while “navigating the evolving financial landscape.

Also Read:RBI Offers Credit Lines Via UPI to Small Finance Banks: What It Means for BorrowersiSpirt Sets Up Initiative for Priority Sector Lending Through OCEN. Here’s How it WorksRBI Launches MuleHunterAI to Tackle India’s Escalating Financial FraudsThe post RBI Deputy Governor Pushes for Stronger AI Governance in Financial Institutions appeared first on MEDIANAMA..