The Reserve Bank of India (RBI) is reportedly developing a centralised Artificial Intelligence-based fraud detection network aimed at monitoring suspicious transactions across UPI, credit cards, and other digital payment platforms in real time. The move comes amid a sharp rise in cyber scams and payment frauds, signalling the regulator’s intent to create an integrated safety grid for India’s growing cashless economy.
Mumbai, October 22 —
India’s financial arteries are going digital, and the Reserve Bank of India wants to ensure they remain secure.
Sources within the central bank have confirmed that the RBI is building an AI-driven fraud detection system to track and flag abnormal payment behaviour across multiple digital channels simultaneously — from Unified Payments Interface (UPI) and IMPS to credit card gateways and e-wallets.
The project, internally code-named “SurakshaNet”, is being designed in collaboration with the National Payments Corporation of India (NPCI), major banks, and cybersecurity agencies. Its goal: to make the RBI a “real-time sentinel” for India’s trillion-dollar digital transaction ecosystem.
“The idea is to detect suspicious activity as it happens, not after losses occur,” said a senior RBI official involved in the project. “With AI and behavioural analytics, we can connect the dots across institutions — something no individual bank can do on its own.”
The Need for a Central Nervous System
Digital transactions in India crossed ₹18 lakh crore monthly through UPI alone in 2025, but the convenience has been shadowed by growing cybercrime.
According to RBI data, reported digital payment frauds rose 36% year-on-year, ranging from phishing scams and QR code tampering to synthetic identity thefts and account takeovers.
Each bank currently runs its own detection algorithms, but criminals often exploit gaps between systems — hopping from one platform to another before an alert triggers.
“What we lack is a national brain that sees all transactions at once,” said Rohit Kumar, Head of Digital Risk at ICICI Bank.
The upcoming AI network aims to fill that gap by integrating anonymised transaction data from banks and payment aggregators into a central monitoring hub.
How SurakshaNet Will Work
The system will employ machine-learning models trained on historical fraud data and behavioural anomaly detection.
For example:
- If an account suddenly initiates hundreds of micro-transactions within minutes,
- or if UPI IDs are accessed from geographically impossible locations,
- or if a merchant terminal’s refund pattern deviates drastically from its average —
the AI engine will raise a “red flag alert” in real time.
The flagged data will then flow to participating banks and enforcement agencies through a secure API, allowing them to freeze, verify, or block suspect accounts immediately.
The network will not store personal details but will operate on tokenised and encrypted transaction identifiers, ensuring privacy compliance under India’s Digital Personal Data Protection Act (2023).
Institutional Collaboration
The pilot, supervised by RBI’s Department of Payment and Settlement Systems (DPSS), involves NPCI, the Indian Computer Emergency Response Team (CERT-In), and four major banks — SBI, HDFC, Axis, and Paytm Payments Bank.
Each participant will provide anonymised data sets for AI training under strict non-disclosure protocols.
Private cybersecurity firms like Tata Elxsi and Quick Heal Technologies are assisting with algorithm testing.
“This is India’s answer to the global Financial Intelligence Unit model — but with AI as its spine,” said Dr. Shailesh Gandhi, former Executive Director, RBI.
The Global Benchmark
Many countries have already adopted similar frameworks:
- The UK’s “HawkEye” system integrates AI alerts across banks.
- Singapore’s MAS uses predictive analytics to flag fraudulent fintech apps.
- U.S. Federal Reserve runs a shared anomaly database with commercial lenders.
The RBI’s initiative seeks to create an Indian equivalent, tailored to high-volume UPI traffic and multilingual consumer interfaces.
Consumer Protection Angle
The RBI has faced criticism for delayed responses to digital frauds, with victims often trapped in bank-to-bank blame loops.
The new AI grid will reduce response time from days to seconds, making reversals or freezes possible before funds vanish into mule accounts.
A 24×7 “Fraud Command Centre” is being planned at the Reserve Bank’s Belapur data hub, staffed by cyber analysts and AI specialists.
“The first 10 minutes after a fraudulent transfer are critical,” said Cyber Crime SP Sanjay Patel. “With automated cross-alerts, we could finally beat scammers at their own speed.”
Legal and Regulatory Framework
The initiative aligns with provisions of the Payment and Settlement Systems Act (2007) and the RBI Master Directions on Digital Payment Security Controls.
A forthcoming notification will make participation mandatory for all regulated entities, including non-bank payment intermediaries.
Additionally, the proposed Digital India Bill will likely classify “AI-enabled financial fraud” as a distinct offence, carrying higher penalties for cross-platform collusion.
AI Ethics and Privacy Concerns
While experts welcome the system’s proactive approach, civil-rights groups urge caution.
“Mass-scale transaction monitoring, even anonymised, risks sliding into financial surveillance if guardrails aren’t clear,” warned Anjali Sharma, policy analyst at the Internet Freedom Foundation.
The RBI has assured that no personal identifiers or chat metadata will be processed; instead, risk models will operate only on numerical and behavioural parameters.
Integration with Law Enforcement
Once live, the platform will directly feed alerts to state cyber cells and the Indian Cyber Crime Coordination Centre (I4C).
Each flagged transaction will carry a unique case ID for traceability and evidentiary use in prosecutions.
This integration is expected to boost conviction rates, which currently hover around 12% for online financial crimes due to lack of standardised digital evidence.
Banks Gear Up
Private banks have already begun scaling up their internal systems to sync with the RBI’s upcoming API.
HDFC Bank has hired over 400 AI analysts, while Axis Bank has partnered with Amazon Web Services (AWS) to host predictive risk models on secure cloud nodes.
“It’s like an immune system — the more data it sees, the better it defends,” said Deepa Verma, Chief Risk Officer at Axis Bank.
Economic Impact
According to RBI estimates, digital frauds cost India’s banking sector ₹1,800 crore in 2024-25.
If the AI network reduces even 20% of such losses, it could save over ₹360 crore annually and restore consumer trust — vital for the ₹250-trillion digital economy goal by 2030.
Fintech start-ups also see opportunity in the rollout, anticipating a new wave of RegTech (regulatory technology)solutions built atop the RBI framework.
Challenges Ahead
- Data-sharing Hesitation: Banks fear competitive exposure.
- Model Bias: AI trained on skewed data might misclassify genuine users.
- Infrastructure: Smaller cooperative banks may struggle with integration costs.
RBI insiders acknowledge these hurdles but argue that phased adoption and subsidies for smaller entities will smooth the transition.
AI and Financial Inclusion
The RBI emphasises that the system is not only about fraud detection but also trust reinforcement.
By making digital payments safer, the network could encourage more rural and elderly citizens to adopt UPI and mobile banking.
A proposed public dashboard will publish anonymised fraud statistics — district-wise and platform-wise — to promote transparency.
Timeline and Next Steps
The pilot phase of SurakshaNet is expected to begin in January 2026, covering 10 million test transactions.
A nationwide rollout could follow by mid-2026, after security audits and performance reviews.
An RBI circular will soon invite stakeholder feedback from fintechs, banks, and consumer organisations.
International Cooperation
The central bank is in talks with the Interpol Global Financial Crime Task Force to share anonymised threat intelligence.
India may also propose an Asian Fintech Security Alliance at the next G20 Digital Ministers’ Meeting, to create interoperable AI fraud standards across borders.
Conclusion: The Algorithm as Watchman
The RBI’s AI-based fraud detection network signals a paradigm shift — from reactive banking security to predictive financial governance.
For India’s 300 million digital-payment users, it could mean safer transactions and faster grievance redressal. For scammers, it marks the end of anonymity.
The central bank’s challenge now lies not in code, but in coordination — linking institutions, ethics, and innovation into one secure digital heartbeat.
#RBI #DigitalPayments #AIFraudDetection #CyberSecurity #UPI #Fintech #BankingInnovation #SarhindTimes
+ There are no comments
Add yours