Winning in the AI Era: The New Playbook for Indian Banks - Inaugural Address by Shri Sanjay Malhotra, Governor, Reserve Bank of India at the FIBAC 2026 Conference, Mumbai, August 11, 2026 -

Distinguished dignitaries, bankers, captains of industry, friends from the technology sector, fintechs, policymakers, press, academia and colleagues from the RBI, good morning to all of you!

It is my pleasure to be back here at the 2026 edition of the FIBAC. It is an important annual conference that brings together the stalwarts of banking and industry to exchange ideas, foster innovation and strengthen collaboration. I congratulate both FICCI and IBA for putting together this event and thank the organisers for giving me the opportunity to share my thoughts today.

The theme of the conference – Artificial Intelligence - has been well chosen. It is apt and timely. It is a theme that, I believe, will define this decade of Indian banking as decisively as liberalisation defined the 1990s and digitalisation defined the 2010s.

I also like the use of the word "playbook". Artificial Intelligence is not a single technology to be procured, nor a project to be completed. It is a new way of doing business, of running a bank. It is a shift in how we evaluate risk, serve customers, price capital, and organise institutions. Many banks in this room are already deploying AI, and many more are considering it. The only question is whether you shape the AI journey with intent, or you let it shape you by default.

Building on FIBAC 2025

Before I turn to AI, let me briefly reconnect with something I shared at this very conference last year. At FIBAC 2025, I had spoken of three priorities that would guide our regulation-making going forward: strengthening financial stability, enhancing ease of doing business, and expanding bank credit while reducing the cost of intermediation.

On strengthening financial stability, we have taken a number of measures. We have finalised the standardised approach for credit risk capital, ECL framework, Effective Interest Rate (EIR) related changes in investment guidelines, prudential norms on project finance, related party transactions, dividends policy, guidelines on Net Open Position (NOP) among others. We are well on target to implement all applicable Basel III guidelines with effect from April 1, 2027 on a calibrated glide path. The regulatory architecture is further bolstered by our enhanced supervision, especially with regard to technology risk.

On enhancing ease of doing business too, we have made good progress. We have reduced regulatory burden on Boards, consolidated regulatory and supervisory instructions, streamlined forms of business, strengthened PRAVAAH, harmonised control and assurance functions, rationalised current-account and working-capital norms, delegated certain foreign exchange related approvals to ADs, etc.

On expanding bank credit, we have continued to strengthen the public digital rails like the Account Aggregator ecosystem and the Unified Lending Interface, that lower the cost of originating and underwriting credit, particularly for MSMEs and underserved borrowers. Rationalisation and updation of regulations related to PSL, project finance, Alternative Investment Funds (AIFs) and acquisition finance, among others, will enhance credit flow to deserving sectors.

Measures such as removal of Investment Fluctuation Reserve (IFR), revised norms on interest rate on deposits and rationalisation of DICGC premium, LCR and CRR shall reduce the cost of intermediation.

We will continue to work on these areas.

Why This Moment Matters

Let me now come to the theme of the conference – AI and start with why the Reserve Bank considers this a matter worthy of our attention.

Every major technological revolution has expanded the frontier of human capability. Steam multiplied muscle-power, electricity multiplied energy, computers multiplied calculation, and the internet multiplied connectivity. The AI wave goes deeper: it multiplies intelligence. AI extends the capability to make judgments at a scale and speed no human workforce could match. That is precisely its promise, and precisely its risk.

India today sits at a unique vantage point. We have the world's most advanced public digital infrastructure – Aadhaar, UPI, Digilocker, ONDC – and more like the Account Aggregator and ULI that are being built on the conviction that infrastructure should be a public good on which private innovation can flourish. AI, layered on top of this stack, has the potential to do for financial judgment what UPI did for financial transactions: make it instant, granular, and available to the last mile. AI, deployed well, can close existing gaps in financial inclusion faster than any preceding generation of technology. Deployed carelessly, it can also entrench new forms of exclusion and instability at a pace regulators and banks may struggle to keep up with.

The Reserve Bank's own Committee for the Framework for Responsible and Ethical Enablement of AI (FREE-AI), which submitted its report last year, put this tension nicely. Let me speak of this tension now.

The Case for AI in Indian Banking

I will be unambiguous: the Reserve Bank sees AI as a capability to be responsibly harnessed and not merely as a risk to be contained. There are at least five reasons why Indian banks cannot afford to sit on the sidelines.

First, AI changes the economics of credit delivery fundamentally. Traditional underwriting relies on financial history – precisely the data that is thin or absent for a new-to-credit borrower, a gig worker, or a small enterprise without formal books. 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.

At the same time, AI-enhanced credit risk models, liquidity forecasting, and scenario analysis allow banks – and, indeed, us, as the regulator – to see emerging stress earlier than lagging financial statements permit.

Second, AI allows banks to serve customers better – provided it is used to augment rather than merely replace human judgment. 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. AI-assisted grievance redressal, and personalised financial guidance can enhance service quality to customers.

Third, and perhaps most important for a country of our size and diversity, AI has the potential to be a profoundly inclusive technology. Voice interfaces in Indian languages can simplify banking by removing the language barrier. Predictive models can identify borrowers on the cusp of default early enough to counsel rather than merely recover. Used well, AI may be the most powerful accelerator to financial inclusion.

Fourth, it can enhance operational efficiency. There is scope to reduce cost to income ratios or intermediation costs in India. Effective adoption of AI can significantly improve the productivity of Indian banks across operations, sales and customer service, and credit and collections. Document processing, reconciliation, and internal audit sampling are all ripe for AI-assisted automation, freeing skilled staff for judgment-intensive work. It can automate transaction reporting, and regulatory return preparation, reducing both compliance cost and the operational risk of manual error.

Fifth, it is AI that can beat AI delivered fraud. Fraud today moves at the speed of an API call. A rules-based fraud engine, however well designed, is perpetually one step behind a fraudster who adapts more frequently. It is only machine-learning models which continuously learn from transaction patterns and can identify anomalies in real time rather than after the loss has crystallised.

This list is illustrative, not exhaustive, and I do not offer it as a mandate. Every bank's playbook should be its own – shaped by its customer base, its risk appetite, and its capacity to govern what it deploys.

But I would urge every bank present here, to ask themselves as to where it stands in AI adoption and how does it accelerate the adoption. You will need to invest in technology: IT infrastructure, talent, skilling and reskilling, forging sustainable partnerships and building governance structures. None of this happens overnight, and none of it happens by accident. It requires a deliberate, board-driven strategy, backed by sustained investment, and strong intent rather than a series of disconnected projects.

The Risks We Must Keep Firmly in View

I now turn to the second half of the playbook, which pertains to the risks.

The first risk is the "black box" problem. Many advanced AI models – particularly deep learning and generative systems – do not readily explain their own reasoning. When an AI system recommends against extending credit to a small business, both the borrower and the regulator are entitled to know why. Opacity is not merely an inconvenience; it strikes at the heart of accountability. It makes it exceedingly difficult for auditors, boards, and the Reserve Bank to be confident that a model is doing what it was designed to do.

The second risk is bias and exclusion. A model trained on historical lending data may, if left unchecked, learn and perpetuate existing biases – biases against certain geographies, certain occupations, certain communities. An algorithm that appears neutral on its face can produce deeply discriminatory outcomes in practice. Fairness in AI-driven finance is not a compliance checkbox; it is a design requirement from day one.

The third risk is concentration and herding. If a handful of foundation models, or a handful of technology vendors, come to underpin credit and trading decisions across much of the banking system, an error, a bias, or a vulnerability in that shared infrastructure ceases to be one bank's problem and becomes a systemic one. AI-driven trading models, if too similar across institutions, can synchronise behaviour in stressed markets and amplify volatility rather than dampen it – a risk this Reserve Bank watches with particular care.

The fourth risk is third-party and vendor dependence. Very few Indian banks, especially smaller ones, will build foundation models in-house. Most will consume AI capability through vendors and technology service providers. This is entirely understandable – but it does not mean that governance can stop at your own walls. Your outsourcing agreements must carry AI-specific accountability: the right to audit, the right to explanation, and a credible exit plan, should a vendor or model need to be replaced.

The fifth risk is data privacy and security. AI systems are hungry for data, and the temptation to feed them more than what is necessary, retain them longer than what is essential, or use them for purposes beyond what the customer consented to, will be constant. Compliance with the Digital Personal Data Protection Act is the floor, not the ceiling, of what customers should expect from their bank.

The sixth risk is cyber and adversarial vulnerability. AI systems can themselves be targets – through data poisoning, model manipulation, or adversarial inputs designed to fool a fraud detector into waving through a fraudulent transaction. As AI becomes more central to your defences, it also becomes a more attractive target for those seeking to defeat it.

And the seventh – perhaps the most important – is the erosion of human judgment and 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. Meaningful human oversight – the ability to explain, to intervene, and, where necessary, to override – must remain a design principle, not an afterthought.

What We Expect, and What You Should Expect of Us

The Reserve Bank's approach to AI, articulated through the FREE-AI Committee's recommendations and draft guidelines on Model Risk Management resolves the tension between the promise and the risk of AI.

It rests on a simple philosophy: innovation and safety are not opposing goals; they are complementary requirements of a durable financial system. We have deliberately chosen a principles-based, proportionate approach over a rigid, prescriptive one, because AI capability and risk will look different for a large bank running proprietary models than for a small bank utilising a vendor's off-the-shelf product.

That said, certain expectations will apply across the board. I would urge every institution here to treat the following as immediate priorities rather than distant compliances:

Maintain a complete inventory of every AI system in use – including those embedded in vendor products – so that neither you nor we are ever surprised by what is running inside your institution.

Establish board-approved AI governance policies, with clear accountability for outcomes, not merely for technology procurement.

Build the capacity to explain AI-driven decisions that materially affect a customer, particularly in lending and fraud outcomes.

Red-team and stress-test AI systems before deployment and periodically thereafter, just as you would stress-test any other material risk.

Preserve meaningful human oversight at every point where an AI system's error could cause material harm to a customer or to financial stability.

We, in turn, are committed to engaging with the industry as this technology and its risks evolve through a willingness to learn alongside you rather than regulate from a distance; and to providing proportional, consultative, evidence-based and agile regulation making and supervision.

We also remain committed to providing the regulatory sandbox as a safe space for testing innovative use cases. We shall continue to facilitate and catalyse development of common utilities such as MuleHunter and the proposed Digital Payments Intelligence Platform to strengthen fraud detection and safeguard the system.

Concluding Thoughts

Let me now conclude.

We have much at stake –building further on the highly successful PMJDY; an MSME credit market, still underserved, estimated at the tens of lakhs of crore rupees, a retail credit culture that is only now maturing, customer service that can be vastly improved, intermediation costs that can be reduced further; and digital frauds that must be curbed.

We need to leverage AI for this. The banks that will win in the AI era will not necessarily be the ones that adopt the most AI, or the fastest. They will be 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 that has always been, and will remain, the true capital of Indian banking.

The role of the bank boards, the risk officers, the technologists, and yes, the regulator is critical in this regard. We all must work together, deliberately, and quickly for this purpose.

I look forward to this journey with you.

I wish the conference much success.

I also wish you all a happy Independence Day in advance.

Thank you.