India's Reserve Bank of India Wants AI in Banking to Approve Loans Humans Would Reject

Reserve Bank of India Pushes AI Adoption in Banking

Reserve Bank of India (RBI) Governor Sanjay Malhotra delivered a landmark speech at the FIBAC conference in Mumbai, positioning AI in banking as "a capability to be responsibly harnessed and not merely as a risk to be contained."1

1

The central bank chief outlined a vision where AI adoption in banking could reshape India's financial landscape, particularly for populations currently excluded from traditional banking services.

Malhotra's remarks signal a significant shift in regulatory thinking about AI's transformative impact on banking. Rather than viewing artificial intelligence primarily through a risk lens, the RBI sees it as essential infrastructure for expanding access to credit and improving operational efficiency across India's banking sector.

AI to Approve Loans Using Alternative Data

The governor's most striking proposal centers on using AI to approve loans for borrowers that human assessors would likely reject. "AI models, trained on alternative data - cash flows, GST filings, utility payments, digital footprints - can extend the frontier of 'bankable' India considerably further than manual underwriting ever could, at a fraction of the marginal cost per loan," Malhotra argued.1

1

This approach targets first-time borrowers, gig workers, and small businesses without formal books—segments traditionally invisible to conventional creditworthiness assessment. By analyzing unconventional data points that manual underwriters struggle to justify, AI for credit access could unlock lending to millions of Indians who remain underbanked or rely on unregulated lenders with high costs.2

2

Expand Financial Inclusion Through AI

Malhotra identified AI as "the most powerful accelerator to financial inclusion" when used properly.1

1

India recognizes 14 major languages spoken by ten million or more residents, plus eight heritage languages. With rural literacy rates below 80 percent, voice interfaces in local languages powered by AI could dramatically expand financial inclusion by making banking accessible to populations previously excluded by language and literacy barriers.5

5

The governor also highlighted AI's capacity for early intervention: "Predictive models can identify borrowers on the cusp of default early enough to counsel rather than merely recover."1

1

This shift from reactive recovery to proactive counseling represents a fundamental rethinking of risk management in banking.

Transform Indian Banking Operations and Customer Service

Beyond lending, Malhotra sees AI reshaping broader banking operations. "A relationship manager assisted by an AI system that presents the right product, the right risk flag, can serve a higher number of customers more efficiently," he noted.1

1

AI-enhanced credit risk models, liquidity forecasting, and scenario analysis could help both banks and regulators spot emerging stress earlier than traditional financial statements permit.

Bank CEOs at the conference highlighted progress in generative AI and agentic AI deployment. "The highest use of AI will be through agentic AI, where we create AI agents for customer service," said Ashok Chandra, MD and CEO of Punjab National Bank.4

4

Standard Chartered's PD Singh reported a 30% improvement in technology implementation speed using AI assistance.4

4

AI Governance and the Black Box Problem

Responsible AI deployment requires addressing what Malhotra called "the black box problem." Many advanced AI models, particularly deep learning and generative systems, don't readily explain their reasoning. "When an AI system recommends against extending credit to a small business, both the borrower and the regulator are entitled to know why," he stated.1

1

Without explainability, auditors, boards, and regulators cannot verify that models function as designed.

The governor emphasized that AI governance must start at the board level. "There has to be a very clear board level AI governance framework in the bank," said Debadatta Chand, CEO of Bank of Baroda.4

4

Banks must maintain the capacity to explain AI-driven decisions affecting customers, particularly in lending and fraud detection outcomes.

Maintaining Human Oversight and Accountability

Malhotra drew a firm line on accountability: "No matter how sophisticated the model, the responsibility for a bank's decisions rests with the bank, not with its algorithm. 'The model decided' can never be an acceptable answer to a customer, an auditor, or the Reserve Bank of India."1

1

Meaningful human oversight—the ability to explain, intervene, and override—must remain a design principle, not an afterthought.

This insistence on human oversight addresses concerns about algorithmic bias against certain geographies, occupations, or communities. Banks must preserve human control at every point where AI errors could cause material harm to customers or financial stability.2

2

Cybersecurity Risks and Industry-Level Response

Hardik Shah, Managing Director at Boston Consulting Group, warned that "cyber risk is a massive potential hurdle" with AI adoption. The cost of creating cyberattacks has fallen 17-fold while attack speed has increased significantly.3

3

Malhotra noted that "it is AI and AI alone that can beat AI delivered frauds."5

5

Shah argued cybersecurity cannot be addressed by individual banks in isolation, calling for "a common utility" with industry-level infrastructure and investment.3

3

He advocated for an AI sandbox similar to India's digital public infrastructure, including the Unified Lending Interface and Account Aggregator framework, with third-party vendor registries and accreditation processes.

Data Quality and Reskilling Challenges

Data quality emerged as a critical challenge for effective AI deployment. Bank executives said access to reliable data and the ability to use it across operations will determine how effectively AI can be deployed.4

4

Banks must also invest in reskilling and upskilling employees to ensure they can use new technology effectively while maintaining necessary controls.

RBI's Five-Point Framework for Banks

Malhotra established clear expectations for responsible AI deployment. Banks must maintain complete inventories of every AI system in use, including vendor-embedded products. They must establish board-approved AI governance policies with clear accountability for outcomes. Banks should build capacity to explain AI-driven decisions, red-team and stress-test AI systems before deployment and periodically thereafter, and preserve meaningful human oversight at every critical decision point.1

1

The governor concluded that winners in the AI era won't be the fastest or heaviest adopters, but "the ones that adopt it with the deepest understanding of what they are deploying, the clearest accountability for its outcomes, and the strongest commitment to the customer's trust."1

1

He stressed this transformation requires "deliberate board-driven strategy backed by sustained investment" and won't happen overnight or by accident.5

5