Part II — Payments

Chapter 14 · Fraud, Mule Accounts and Dispute Resolution

UPI's cryptography has not been broken — its customers have been talked out of their money. This chapter walks the dark side of the rails: how fraud is actually classified, how mule accounts launder the proceeds in a race against the freeze, the RBI's customer-liability ladder stated clause by clause, the reversal clocks for failed payments, and the dispute route that ends at the Integrated Ombudsman.

📖 Chapter 14 of 35 ⏱ ~60 min read 🎓 Session 17 · CILO 3 Core 60% Stable 40% Volatile 0% 📝 Quiz · 30 MCQs

🎯 Learning Objectives

1. How Payment Fraud Actually Works

Chapter 9's warning is where every fraud story now begins: UPI's authentication — device binding plus PIN — is strong, so fraudsters stopped attacking the cryptography and started attacking the person holding it. Classify what they do into three families, because the families carry different liabilities and remedies.

Authorised-push-payment (APP) fraud: the customer initiates and authenticates the payment, deceived about its purpose. The fake collect request that a flustered user approves; the screen-sharing app that lets a "bank officer" watch and coach; the "digital arrest" that keeps a victim on the line while payments flow; the task-and-investment scam that pays small rewards before taking the big sum. Technically impeccable transactions — every hop of Chapter 9's flow executed correctly, with the human as the broken component. Account takeover: the fraudster gains control — SIM-swap to capture OTPs, stolen credentials, malware — and transacts without the customer. These are genuinely unauthorised transactions, and the liability machinery of Section 3 was built for them. System-side compromise: breach or deficiency at a bank, processor or merchant — the rarest family, and the one where the bank's own liability is clearest.

📝
The classification question to ask of any case: who authorised the payment? If the customer did — tricked or not — it is APP fraud, chargebacks do not exist (Chapter 13), and recovery depends on the freeze race of Section 2. If nobody authorised it, the RBI's liability ladder applies in full.

2. The Mule Economy

Money leaves the victim's account and must land somewhere. It lands in a mule account — an ordinary bank account whose holder receives fraud proceeds and passes them on, taking a small cut. The mule layer is the laundering: proceeds hop across multiple accounts in hours, splitting and re-splitting, until they surface as cash, crypto or goods far from the crime.

Mules are recruited, not born. "Earn from home — use your bank account" advertisements target students and first-time account holders; some mules know exactly what they are doing, others genuinely believe they are doing "payment processing" for a business. The law is less interested in the belief: renting out your account is money laundering, and the account, the credit record and the liberty are all on the line. For the fraudster, mules provide the essential service — clean accounts with clean histories that no velocity filter flags on day one.

Now the race that Chapter 9's irrevocability created. An authorised UPI payment cannot be reversed by rule, so recovery means freezing the money while it still sits in a mule account, before the cash-out. India built the machinery for exactly that race: the 1930 cybercrime helpline and the national reporting system behind it, which let a victim alert banks and enforcement within minutes, so accounts down the chain can be frozen before withdrawal. Every minute the victim waits is a minute the money moves — which is why this chapter's first-hour checklist exists, and why the liability ladder rewards fast reporting so heavily.

📋 Stable content — Reviewed: October 2026 · Next review: April 2027

3. The Customer-Liability Ladder

The governing framework is the RBI's circular of 6 July 2017 — Customer Protection: Limiting Liability of Customers in Unauthorised Electronic Banking Transactions — stated here as it reads.

Zero liability arises in two cases. First, where the fraud happened through contributory fraud, negligence or deficiency on the part of the bank — irrespective of whether the customer reported anything. Second, where the breach lies with neither bank nor customer (a third-party breach), and the customer notifies the bank within three working days of receiving the bank's communication (the SMS, the alert) about the transaction.

Customer negligence — the credential-sharing case — inverts the burden: the customer bears the entire loss until the transaction is reported; everything after reporting is the bank's. Report in four to seven working days (third-party breach) and liability is capped at the transaction value or the table amount, whichever is lower: ₹5,000 for BSBD (basic savings) accounts; ₹10,000 for other savings accounts, PPIs, MSME current accounts and credit cards up to ₹5 lakh; ₹25,000 for the larger current/overdraft accounts and cards above ₹5 lakh. Beyond seven working days, the bank's Board-approved policy decides.

Two clocks complete the ladder. On notification, the bank must make a shadow reversal — credit the customer within 10 working days, value-dated to the transaction date, without waiting for insurance or investigation to finish. And the complaint must be resolved with liability established within 90 days; if the bank cannot resolve it in time, it pays the compensation anyway. Banks must also give customers 24×7 reporting channels, reply-able SMS alerts, and an acknowledged complaint number — because the timestamps decide everything.

📝
Recent addition: RBI has since issued directions on compensation for small-value fraudulent transactions, including a one-time reimbursement facility of up to ₹50,000 for eligible victims of digital payment fraud — a widening of the safety net beyond the 2017 ladder. Watch this space at review time; consumer-protection rules are the most actively updated pages in this book.
📋 Stable content — Reviewed: October 2026 · Next review: April 2027

4. Failed Transactions and the Reversal Clocks

Chapter 9 diagnosed the "money debited, merchant says not received" failure — the credit leg broke. Chapter 13 priced the wait. This section gives it its remedy: under the RBI's turnaround-time (TAT) harmonisation framework of 2019, every payment rail carries a prescribed timeline within which a failed transaction must be auto-reversed to the customer's account — and if the institution misses the timeline, it pays compensation of ₹100 for every day of delay, credited automatically.

The design principle matters more than the per-rail table: the customer should never have to ask for her money back. Auto-reversal is the system's job; the compensation is automatic too — not a claim to be argued, a credit that arrives. When a bank instead demands a written application for a failed-transaction reversal, it is not following the framework; and the dispute route of Section 5 exists for exactly that behaviour. Keep the distinction from fraud clear: a failed transaction is a system event with a clocked remedy; a fraud is a crime with a liability ladder. Mixing the two confuses the remedy.

5. The Dispute Route, End to End

Every remedy in this chapter ends in the same escalation ladder. Learn it once, use it for anything.

  1. Report immediately. The bank's channels (app, helpline, reply-to-SMS) — and, for fraud, 1930 and the cybercrime portal in parallel. Get the complaint number; the clock is now running.
  2. The bank responds. Shadow reversal within 10 working days where liability rules require it; resolution within 90 days at the outside.
  3. Escalate to the bank's internal ombudsman / grievance head if the answer is wrong or absent.
  4. Thirty days from the bank's reply — or from the complaint, if it never replies — approach the RBI Integrated Ombudsman, the one-nation-one-ombudsman system of 2021 that merged the separate banking, NBFC and payment ombudsman schemes. Filing is online, free, and jurisdiction is centralised.

What the ombudsman can do — direct reversal where the ladder required it, compensation for delay, and the consequential interest — is why the evidence discipline of the fraud walkthrough matters: screenshots with timestamps, the SMS trail, complaint numbers, and the dates of every communication decide ombudsman cases. For card disputes, the chargeback machinery of Chapter 13 runs in parallel with this ladder; for UPI there is no chargeback, and this ladder is the whole route.

💡
Pro Tip — the first hour after fraud: (1) call 1930; (2) call the bank and get a complaint number; (3) freeze the channel — block the card, change the PIN, get the SIM reissued if swapped; (4) screenshot everything before it scrolls away; (5) write down the timeline while it is fresh. Speed is the only variable the victim controls, and the liability ladder prices it in days.

6. Fraud as a System Problem

Individual cases run on the ladder; the system runs on the statistics, and the honest reading of them is double-sided. By value, reported digital payment fraud is a small fraction of the trillions of rupees that move; the rail is not lawless. But by victim count and by fear, it is large enough to shape behaviour — and APP fraud in particular survives every technical control, because its broken component is a person the system cannot patch.

So the prevention stack is layered. At the rail: device binding, transaction limits, cooling periods for new beneficiaries. At the banks: velocity and mule-account detection — a savings account suddenly routing hundreds of transfers is a signature no fraudster enjoys. At the ecosystem: the 1930 freeze-race machinery, merchant-verification labels on collect requests, and education campaigns that teach the one sentence every user should know: no real bank or official will ever ask for your PIN, OTP or a screen-share. Notice that this, too, is Chapter 5's thesis under stress: public infrastructure is measured by its failure modes, and the fraud stack is the failure mode of the payments stack.

One more layer remains before Part II closes: the border. Fraud follows money across jurisdictions, payments increasingly do too, and the RBI's Payments Vision 2028 carries both — the international linkages of UPI and the risk framework around them. Chapter 15 takes the rail abroad.

🛠 Hands-On: The Fraud Case Walkthrough

This is the session's classroom activity, made self-paced. Four cases; for each, classify the fraud, run the first hour, compute liability under the ladder, and state the escalation route. Show the rule and the arithmetic together.

Steps

  1. The fake collect request. Asha approves a ₹40,000 collect request during a "refund verification" call. Classify the fraud. Is the liability ladder's zero-liability window her friend here? What should her first hour look like, and what is her realistic recovery path?
  2. The SIM swap. Ravi's SIM is swapped; ₹85,000 leaves his regular savings account in transactions he never made. He reports (a) on working day 2 after the SMS, (b) on working day 5. Compute his liability in each case.
  3. The failed payment. A ₹2,500 UPI payment is debited but never reaches the merchant; the bank's reversal lands 12 days late. What does the TAT framework owe her, beyond the ₹2,500?
  4. The rented account. A student earns ₹5,000 a month letting a "digital marketing agency" route payments through his account. What is the account, what is his legal position, and what should he do today?
  5. Synthesise. In 120 words: why does the liability ladder price reporting speed so steeply, and what single design change to UPI would most reduce APP-fraud losses — at what cost?
View Solution / Walkthrough

1. APP fraud — Asha authorised the payment with her own PIN, deceived about its purpose. The zero-liability window helps less than it looks: banks argue deception-with-OTP approaches negligence, and there is no chargeback on UPI. First hour: 1930, bank complaint number, screenshots of the collect request and the caller's number, timeline notes. Recovery path: the freeze race — if the ₹40,000 sits in a mule account when the alert lands, it can be frozen and returned through enforcement; if it has been layered out, recovery is uncertain regardless of the ladder. Case 1 is the chapter's hardest lesson: authorised is the fraudster's favourite word.

2. Account takeover — genuinely unauthorised; the ladder applies cleanly. (a) Day 2 is within three working days of the SMS: zero liability; shadow reversal within 10 working days, value-dated. (b) Day 5 is the four-to-seven window: capped at the table amount for a regular savings account — ₹10,000 — or the transaction value if lower; ₹85,000 > ₹10,000, so Ravi bears ₹10,000 and the bank ₹75,000. Same crime, ₹10,000 of difference, three working days apart.

3. A failed transaction, not a fraud: the credit leg broke (Chapter 9's hop 5). The ₹2,500 auto-reverses within the rail's TAT; the 12-day delay owes ₹100 per day of delay beyond the timeline — compensation credited automatically, not applied for. If the bank resists, the dispute ladder ends at the Integrated Ombudsman.

4. His account is a mule account; he is participating in money laundering, knowingly or not, and "they said it was marketing" is not a defence that survives the first notice. Today's actions: stop, report to the bank and 1930 with the recruiter's messages preserved, and expect the account to be frozen — the freeze that protects fraud victims is the same freeze that ends the mule's banking.

5. Strong answers: the ladder prices speed because recovery is a race — money that sits in a mule account can be frozen, money layered out cannot, so fast reporting is the only lever that changes the outcome. Design-change candidates and costs: a default cooling period or a delay on new-beneficiary large transfers (cost: friction in genuine commerce — Chapter 9's collect-request elegance deliberately traded speed for convenience); better labels and friction on collect requests above a threshold; mandatory cooling on first-time high-value payees. Full credit for naming any one and pricing its cost honestly.

Key Takeaways

1

Classify first: who authorised the payment? APP fraud (the customer did, deceived), account takeover (nobody did), system compromise (the bank's fault).

2

Mule accounts launder the proceeds; recovery is a race between the 1930 freeze and the cash-out. Irrevocability made speed everything.

3

The 2017 ladder: report within 3 working days — zero liability; 4–7 days — capped ₹5,000/₹10,000/₹25,000 by account; negligence — the customer pays until reported; shadow reversal in 10 working days.

4

Failed transactions are not frauds: TAT auto-reversal with ₹100-a-day delay compensation — money returns without being asked for.

5

The dispute ladder: bank → complaint number → 30 days → RBI Integrated Ombudsman (2021). Evidence with timestamps decides ombudsman cases.

Test Your Understanding

Choose one best answer per question. A calculator is allowed for the liability computations; the ladder of Section 3 is the rulebook. Aim for 24/30 before moving to Chapter 15.

1. A user approves a fake collect request during a "refund verification" call. This is:

2. Transfers made after a SIM swap, with no customer involvement, are classified as:

3. Payment fraud in the UPI era targets the customer because:

4. A mule account is one that:

5. The legal position of a student who rents out his account to a "marketing agency":

6. The national cybercrime helpline number for financial fraud is:

7. Zero liability for a third-party breach requires reporting within:

8. If the fraud arose from the bank's own deficiency, the customer's liability is:

9. Where the customer shared her OTP and the account was drained, the loss before she reports is borne by:

10. Reporting a third-party breach on working day 5 caps a regular savings customer's liability at: (Calculator allowed.)

11. A ₹7,000 third-party-breach loss, reported on day 5, on a regular savings account: the customer's liability is: (Calculator allowed.)

12. If the customer reports the breach after seven working days, liability is set by:

13. The shadow-reversal clock requires the bank to credit the customer within:

14. If the bank cannot resolve the complaint and establish liability within 90 days:

15. The recent addition to the compensation regime is:

16. Under the TAT harmonisation framework, a bank that reverses a failed payment beyond the timeline pays:

17. A "debited but not received" payment is best treated as:

18. The one-nation-one-ombudsman system of 2021 merged the schemes covering:

19. The customer may approach the Integrated Ombudsman after:

20. What decides most ombudsman cases in the customer's favour is:

21. The correct first hour after discovering a fraud contains, in order:

22. Ravi reports an ₹85,000 SIM-swap fraud on working day 2 after the alert. His liability is: (Calculator allowed.)

23. The same ₹85,000 fraud reported on working day 5 leaves Ravi liable for: (Calculator allowed.)

24. The detection tool most aimed at the mule layer specifically is:

25. The single sentence that defeats most APP-fraud scripts is:

26. The 2017 framework obliges banks to provide:

27. The design tension in adding friction (cooling periods, delays) to fight APP fraud is:

28. For dispute remedies, cards and UPI differ because:

29. The chapter's one-line summary of the whole machinery:

30. Part II's final chapter takes the payments story: