mobile-logo
RBI_banner
14 Jan

RBI Digital Payments Intelligence Platform (DPIP)

by Shailendra Gupta
in Blog
Comments

Attempt to compile publicly available Info at one point, with the primary intention of raising awareness about this important RBI initiative.

1. LAUNCH TIMELINE

Official Announcement

  • Announced: October 1, 2025 (RBI Monetary Policy Statement)
  • Current Status: Under development/pilot phase
  • Pilot Phase: Started early 2025 with 5 banks
  • Scaling: Now expanded to 12+ public sector banks (as of October 2025)
  • Expected Full Launch: Timeline not yet officially announced, but infrastructure is being built progressively. It should be very soon.

Key Milestones

  • Early 2025: RBI Innovation Hub developed initial prototype
  • April 2025: Platform mentioned in public discourse
  • October 2025: Deputy Governor T. Rabi Sankar publicly detailed the platform at the Global Fintech Festival, Mumbai
  • December 2025: State Bank of India and other PSBs received RBI approval for Digital Payments Intelligence Company (IDPIC) – a Section 8 company to operate DPIP

Source: Economic Times, October 14, 2025; BIS Speech by T. Rabi Sankar, October 13, 2025

2. TECHNICAL DETAILS

Phase 1 (Currently Operational): Negative Registry System

  • Integrates data from: Telecom operators + Indian Cyber Crime Coordination Centre (I4C) + Banking fraud databases
  • Flags suspicious or fraudulent entities
  • Real-time information sharing between banks
  • Pattern recognition across multiple institutions

Phase 2 (Rolling Out Soon): AI-Driven Risk Scoring

  • Real-time transaction analysis
  • Instant risk score assignment for each payment
  • Pre-transaction alerts for high-risk payments
  • Enables banks to take preventive actions like – Enhanced due diligence, Additional verification, Temporary debit freezes, and Transaction blocking

Data Sources

  • Mule account databases
  • Telecom data
  • Geographical location data
  • Transaction patterns
  • Historical fraud data
  • Cross-bank intelligence

AI Capabilities

  • Machine learning models trained on fraud patterns
  • Real-time analysis of payment flows
  • Behavioral analytics
  • Anomaly detection
  • Pattern matching across institutions

Source: Economic Times article by Alekh Angre & Saloni Shukla, October 14, 2025

3. RBI’S STRATEGIC RATIONALE

Problem Being Addressed: Fraud Scale in India

  • FY25: 13,516 cases of card and internet fraud
  • Total amount: ₹520 crore in losses. Probably way more because this may be hugely under-reported.
  • Most frauds occur through digital channels (cards, internet banking, UPI)
  • Private sector banks account for the majority of digital frauds
  • Public sector banks report most loan-related frauds

Strategic Objectives

  1. Real-time fraud prevention (not just detection)
  2. Cross-institutional intelligence sharing (breaking data silos)
  3. Proactive risk management (stopping fraud before it happens)
  4. Standardised fraud classification across the banking system
  5. Leveraging AI for scale (billions of transactions need automated intelligence)

RBI Deputy Governor’s Vision

T. Rabi Sankar stated: “The basic idea is to collect information from multiple sources—mule accounts, telecom, geographical location, and more—and train an AI system on this data. The system will generate pre-transaction alerts if it identifies a risk, allowing banks or customers to decide whether to proceed.”

 

Key Philosophy: Responsible AI implementation – balancing innovation with financial stability

Source: BIS Speech, October 13, 2025; RBI Monetary Policy Statement, October 1, 2025

4. WHAT FINTECH COMPANIES CAN EXPECT

For Payment Service Providers (PSPs)

  1. Mandatory Integration: All PSPs will need to connect to DPIP
  2. Real-time Risk Scores: Every transaction will receive a fraud risk score from DPIP
  3. Decision Support: Risk scores supplement (not replace) internal fraud controls
  4. Data Sharing Requirements: Must share fraud-related data with the platform
  5. Enhanced Due Diligence: May need to implement additional verification for high-risk transactions

For Digital Payment Companies

  1. Improved Fraud Detection: Access to cross-industry fraud intelligence
  2. Reduced False Positives: Better risk assessment means less number of legitimate transactions blocked
  3. Compliance Support: Centralised platform helps meet RBI fraud prevention requirements
  4. Cost Savings: Shared infrastructure reduces individual fraud management costs
  5. Faster Response: Real-time alerts enable immediate action

Implementation Requirements

  • Technical integration with DPIP APIs
  • Data sharing protocols
  • Internal process changes for risk-based authentication
  • Staff training on new alert systems
  • Compliance with data privacy standards

Source: Economic Times, October 14, 2025

5. WHAT THE GENERAL PUBLIC CAN EXPECT

Benefits for Consumers

  1. Safer Transactions: AI catches fraud before money leaves the account
  2. Pre-transaction Warnings: Alerts if payment appears risky
  3. Reduced Fraud Losses: Better prevention = fewer victims
  4. Faster Resolution: Coordinated response across banks
  5. Maintained Privacy: System uses pseudonymization and encryption

Potential Friction Points

  1. Additional Verification: High-risk transactions may require extra authentication
  2. Temporary Holds: Suspicious payments may be delayed for verification
  3. False Positives: Some legitimate transactions might be flagged initially
  4. Learning Curve: System improves over time as AI learns patterns

Consumer Protection

  • RBI emphasises responsible AI deployment
  • Explainability tools to understand why transactions are flagged
  • The customer can override or provide additional information
  • Privacy safeguards built into system design

Source: BIS Speech by T. Rabi Sankar, October 13, 2025

6. GLOBAL COMPARISONS

 

Similar Initiatives Worldwide

1. Federal Reserve – FraudClassifier Model (USA)

Launched: June 2020 Type: Classification framework (not AI-driven real-time system)

  • Standardises fraud classification across payment types
  • Helps industry speak “same language” about fraud
  • Voluntary adoption by banks
  • Focuses on ACH, wire, and check payments
  • Integrated with ScamClassifier model (launched 2024)

Key Difference from DPIP:

  • FraudClassifier is a classification taxonomy, not a real-time AI detection system
  • DPIP is more advanced – provides pre-transaction risk scoring

Source: Federal Reserve, FedPayments Improvement, 2020-2025

2. European Central Bank – Digital Euro Fraud Prevention (EU)

Announced: October 2, 2025 Partner: Feedzai (AI fintech company) + PwC Contract Value: €79.1 million (up to €237.3 million)

  • AI-powered fraud risk scoring for digital euro transactions
  • Central fraud detection for 440 million citizens
  • Real-time risk assessment for P2P and P2M payments
  • PSPs use risk scores alongside own controls

Similarity to DPIP:

  • Both use AI for real-time fraud risk scoring
  • Both provide central infrastructure for a distributed payment system
  • Both focus on pre-transaction prevention
  • Both balance innovation with privacy/security

Key Difference with DPIP:

  • The ECB system is for the future digital euro (not yet launched)
  • DPIP is for existing payment systems (UPI, IMPS, cards, and internet banking)
  • RBI is building in-house through their Innovation Hub, whereas ECB outsourced to a private vendor

Source: Feedzai Press Release, October 2, 2025

3. SWIFT – Cross-Border Fraud Detection (Global)

Announced: September 2025 Partners: 13 international banks

  • AI innovation for cross-border payment fraud
  • Doubled real-time fraud detection in trials (10 million test transactions)
  • Privacy-enhancing technologies
  • Cross-border collaboration blueprint

Similarity to DPIP:

  • Multi-institution data sharing
  • Real-time AI-driven detection
  • Focus on collaboration

Key Difference with DPIP:

  • SWIFT focuses on cross-border payments
  • DPIP focuses on domestic digital payments in India

Source: SWIFT Press Release, September 2025

7. RBI’S THOUGHT LEADERSHIP

 

Why DPIP Shows RBI Innovation

1. Proactive, Not Reactive

  • Most central banks focus on post-fraud analysis
  • RBI is building a pre-transaction prevention system
  • Shift from “detect and respond” to “predict and prevent”

2. Collaborative Infrastructure

  • Breaking down data silos between banks
  • Creating a shared intelligence platform
  • Industry-wide benefit from collective learning

3. In-House AI Development

  • RBI’s Innovation Hub is building the platform itself and not outsourcing to private vendors.
  • Maintains complete control over critical financial infrastructure
  • Demonstrates the technical capability of the Indian central bank

4. Scale-Appropriate Solution for Usage of AI

  • India processes billions of digital transactions monthly
  • 97.6% of payments are now digital (RBI data, 2025) – Of course, ignore the black economy, which is another conversation.
  • Traditional fraud detection can’t scale
  • Building an AI-driven platform is necessary for such transaction volume

5. Responsible AI Framework

  • T. Rabi Sankar emphasised a Responsible AI approach
  • Stress-testing AI models
  • Red-teaming to identify vulnerabilities
  • Explainability tools
  • Layered oversight
  • Privacy safeguards

Global Recognition

  • One of the first central banks to build real-time AI fraud prevention at a national scale
  • More advanced than the US Federal Reserve’s classification approach
  • Similar timeline to ECB’s digital euro fraud system (but for live payments)
  • Positions India as a leader in digital payment security innovation

Source: BIS Speech, October 13, 2025; Economic Times, October 14, 2025

8. INSTITUTIONAL STRUCTURE

Digital Payments Intelligence Company (IDPIC)

Incorporation: October 2025 Structure: Section 8 company (not-for-profit)

  • Authorised capital: ₹500 crore
  • Paid-up capital: ₹200 crore Ownership: All 12 public sector banks
  • The function is to operate the DPIP platform
  • RBI’s Role is providing oversight and regulatory guidance

Source: ScanX Trade, December 11, 2025

9. FRAUD CONTEXT IN INDIA

Current Fraud Landscape

  • 46% surge in banking cyberattacks (2025 vs 2024)
  • Shadow AI risks: Employees using ChatGPT, potentially leaking sensitive data
  • AI hallucinations: Chatbots providing wrong information, leading to financial losses
  • Social engineering: Increasing sophistication of scams, especially using AI
  • Mule accounts: Networks of accounts used to move fraudulent funds

Why DPIP is Needed Now

  • Digital payment volume is growing exponentially
  • Fraud techniques evolving faster than traditional detection
  • Cross-bank coordination has been weak
  • Real-time payments (UPI) require real-time fraud prevention
  • International fraud networks operating in India

Source: CISO Forum, 2025; Economic Times, 2025

10. SOURCES IF YOU WANT TO LEARN MORE YOURSELF

  1. Economic Times – DPIP: RBI’s AI tool to detect fraud in real time (October 14, 2025)
  2. Bank for International Settlements (BIS) – Speech by T. Rabi Sankar, Deputy Governor RBI (October 13, 2025)
  3. Federal Reserve – FraudClassifier Model documentation (2020-2025)
  4. Feedzai Press Release – ECB Digital Euro partnership (October 2, 2025)
  5. SWIFT – Cross-border fraud detection announcement (September 2025)
  6. RBI Monetary Policy Statement – October 1, 2025
  7. ScanX Trade – IDPIC incorporation news (December 11, 2025)
  8. Reuters – ECB AI fraud prevention (October 2, 2025)
  9. RBI Innovation Hub

Source: CISO Forum, 2025; Economic Times, 2025

MY TAKE: This is critical infrastructure. Most people won’t notice it. But it matters.

What’s your take on this?

  • If you’re in fintech or banking, are you already working on DPIP integration?
  • If you’re a consumer, would you be okay with occasional extra verification if it meant better fraud prevention and protection?

Share this post

About the Author

Shailendra

Shailendra Gupta
(Co-Founder and CEO of Mind IT Systems)

 

Shailendra is Co-Founder and CEO of Mind IT Systems and is responsible for strategy and business relations.

With around two decades of experience in getting things done in marketing, sales, strategy, delivery, or technology, he has a successful track record of leading startups and mid-size companies and being a prime contributor to stakeholder management, growth, and value creation. A thought leader in the geo-social space, he is highly respected for realizing new paradigms in marketing, solutions, and approaches.