Mumbai: Financial fraud is no longer confined to individual banks or financial institutions, prompting the banking and financial services industry to rethink how risks are identified and controlled. Cybercriminals are increasingly moving money across bank accounts, digital wallets, fintech platforms and multiple digital applications, making institution-level monitoring insufficient to track the full chain of suspicious activity.
Senior risk executives at the eighth edition of the ETBFSI CXO Conclave 2026 in Mumbai said the next phase of fraud prevention would require industry-wide intelligence sharing, common data infrastructure, real-time monitoring and coordinated controls. They said financial institutions need to identify suspicious customers, mule accounts, beneficiaries, devices and transaction patterns across institutional boundaries rather than examining individual transactions in isolation.
Arun Pandey, CGM RM-II at State Bank of India, said organised fraudsters do not operate within the boundaries of individual financial institutions. Money can move from a bank account to a wallet, through another application and then into accounts maintained with multiple institutions. Such movement makes it difficult for any single bank to develop a complete picture of the fraud chain.
Pandey stressed that the industry needs a coordinated approach in which banks can access and share relevant intelligence about suspected fraudsters, mule accounts and beneficiaries. Financial institutions are consequently exploring shared suspect registries and intelligence platforms that could identify risks during customer onboarding and eventually at the transaction level.
Ramanuj Sharma, GM and Deputy Risk Officer at Bank of Baroda, said fraudsters are increasingly targeting people rather than simply exploiting weaknesses in technology. Investment scams, digital arrest fraud and other forms of social engineering allow criminals to manipulate victims over extended periods.
According to Sharma, the response must combine artificial intelligence-based detection with customer awareness and intelligence sharing across financial institutions. He said the larger question is no longer whether risk management is coordinated within an individual bank, but whether the financial industry itself is sufficiently connected to respond to fraud.
Risk executives also highlighted the importance of detecting suspicious activity before money leaves an account. Bharan Kumar Guntupalli, EVP-Risk at NaBFID, said the effectiveness of fraud controls should be measured by how quickly suspicious activity is identified rather than simply by the number of fraud cases detected.
He also cautioned institutions against abandoning traditional fraud controls while focusing on artificial intelligence, cyber risks and digital attacks. Weak validation, inadequate basic controls and overlooked lending-fraud patterns can continue to create vulnerabilities even as newer technologies are deployed.
Zeenat Hamirani, CRO at L&T Finance, pointed to the speed gap between criminals and financial institutions. Fraudsters can change their methods quickly, while institutions need governance, testing and approval before modifying their systems. She said completely eliminating friction from customer journeys is therefore unrealistic. Limited additional verification at high-risk moments can help protect customers without unnecessarily disrupting legitimate transactions.
Anand Viswanathan, Group Chief Risk Officer at Axis Bank, argued that financial institutions should assess risks according to their outcomes rather than treating credit, operational, cyber and technology risks as completely separate categories. A fraud event may originate from a credit decision, process failure, technology vulnerability, inadequate collateral monitoring or an information-security lapse.
Manish Agrawal, Senior EVP-Credit Intelligence & Control at HDFC Bank, said institutions often have substantial amounts of relevant data, but the information remains divided among different systems and functions. Fraud alerts, anti-money laundering signals, mule-account intelligence, onboarding information, transaction monitoring and law-enforcement inputs need to be connected through a common architecture.
Dr Radhakrishna B, Director-Customer Advisory at SAS India, said no single model can solve the fraud problem. Institutions need to combine business rules, anomaly detection, artificial intelligence, machine learning and network analytics. Risk signals associated with customers, accounts, beneficiaries, devices, merchants, ATMs and locations can then be assessed together before a transaction is approved, challenged or blocked.
Amol Padhye, CRO at AU Small Finance Bank, also highlighted emerging risks involving AI-generated attacks, deepfakes, model risk and automated decision-making. Experts said the future fraud-defence framework will require banks, fintech companies, payment platforms, regulators, telecom operators and law-enforcement agencies to share relevant intelligence quickly. The objective is to preserve convenience for legitimate customers while applying the right level of verification and intervention when risk signals emerge.
About the author — Suvedita Nath is a science student with a growing interest in cybercrime and digital safety. She writes on online activity, cyber threats, and technology-driven risks. Her work focuses on clarity, accuracy, and public awareness.
