‘AI vs AI’ in Banks: Why Real-Time Fraud Checks Are Becoming Critical

Banks may soon face a new generation of financial fraud in which artificial intelligence systems are used not only to detect suspicious activity but also to power emerging threats, SBI Chairman CS Setty has warned.

He said lenders will need real-time risk assessment combining identity, behavioural and transactional intelligence as AI adoption expands across financial services.

What Does ‘AI vs AI’ Mean for Banks?

Speaking at the Global Fintech Fest 2026, Setty said banks could find themselves operating in an “AI versus AI” environment, where technology is deployed on both sides of the financial security challenge.

The next major challenge may therefore go beyond using artificial intelligence to automate lending or identify fraud. Banks will also have to defend against AI-powered threats capable of challenging existing fraud-control systems.

Setty stressed that financial services require a particularly high standard of reliability because mistakes can directly affect customer funds and financial security.

Proposal for Conducting Cyber Crisis Drill, Tabletop Exercise (TTEx) & CCMP Readiness Exercise

Why Is 99% AI Accuracy Not Enough in Banking?

Setty said AI systems in financial services cannot operate with a 99% accuracy threshold when an error could have consequences for customers’ money and financial security.

The warning highlights the difference between using AI for ordinary automated tasks and relying on it for sensitive financial decisions. Even a relatively small error rate can become significant when technology is involved in decisions affecting transactions, fraud detection and customer accounts.

As banks expand their use of AI, accuracy therefore has to be considered alongside the financial consequences of incorrect decisions.

How Could Agentic AI Change Financial Services?

The next phase could involve agentic AI, which Setty said may allow technology to act on behalf of customers and institutions instead of simply supplying information or recommendations.

He described the real promise of agentic AI as demonstrating “financial intelligence” and highlighted its potential to democratise access to financial decision-making.

Such systems could give AI a more active role in financial processes. This also increases the importance of safeguards because technology acting on behalf of a customer or institution could have a more direct effect on financial decisions.

Why Will Banks Need Real-Time Fraud Checks?

Setty said an AI-versus-AI environment will require lenders to combine identity, behavioural and transactional intelligence for real-time risk assessment.

This means fraud detection cannot depend on a single signal. Banks will need to assess who is involved in a transaction, whether their behaviour appears consistent and whether the transaction itself shows signs of unusual activity.

Real-time assessment becomes particularly important when AI-powered threats can operate quickly. Detecting suspicious activity after a transaction has already progressed may not provide the same protection as identifying risk while the financial activity is taking place.

What Does the AI Fraud Warning Mean for Customers?

The warning points to a financial environment in which both banking security systems and emerging threats can become more technologically sophisticated. Customers may increasingly interact with AI-supported banking systems while financial institutions use the same broad technology to identify fraud and manage risk.

Setty’s comments also underline why automated systems cannot be treated as infallible. Banks need strong accuracy standards because incorrect decisions can involve customer money and financial security.

As AI systems become capable of taking more active roles, real-time identity, behavioural and transaction checks will become increasingly important to distinguish genuine financial activity from suspicious behaviour.

How Can Customers Stay Safe as AI-Powered Fraud Grows?

The screenshot does not provide specific customer safety instructions from SBI or Setty. However, his warning makes one point clear: financial institutions will need stronger, real-time checks as AI becomes more deeply involved in banking and emerging fraud risks.

For customers, the development means bank security warnings and verification checks should be taken seriously rather than viewed simply as delays in a transaction. As banks move towards systems that assess identity, behaviour and transactions together, such checks will form an important part of protecting customer funds. At the institutional level, Setty said the industry will also need more affordable computing and cybersecurity infrastructure as AI adoption expands.

About the author — Ayesha Aayat writes on cybercrime, digital safety, and emerging online threats. Her work focuses on public awareness, legal clarity, and technology-driven risks.