- Police allege attackers reached three PayU cloud servers and used its payout API for 265 unauthorised transfers worth Rs 25.67 crore. How they got in is not known.
- The case raises a question about who can add a payee to an automated payout system, and who checks that change before money moves.
- Leaders can ask when a new payee triggers a human review, and how fast a cloud provider's alert reaches their own team.
The money left through a door built for machines
When a payments company sends money out, software does the work. A payout API is the channel that software uses to tell a bank to pay a given amount to a given account. No clerk approves each transfer.
That is why the alleged breach at PayU Payments Private Limited deserves attention beyond one company. It raises a question this publication thinks every payer should ask. When software can pay anyone on its list, who controls the list?
What police say happened
MediaNama, citing a Hindustan Times report, says Gurugram Cyber Police arrested two men in Ludhiana on November 15. PayU had filed its complaint on November 9.
Police allege the attackers reached three of PayU's cloud servers and tampered with its payout API. The reported toll is 265 unauthorised transfers touching 156 bank accounts, worth Rs 25.67 crore in total. Police also say the attackers added new payees and sent money without PayU's approval.
These are allegations. Police are waiting for an audit report that should explain how the systems were breached.
The timing is part of the story. MediaNama reports that PayU's cloud service provider flagged suspicious access. PayU's own security team noticed odd API activity at about 11:30 pm on November 5. By then the transfers had already happened. Police have frozen about Rs 1.5 crore.
How a payout attack works
A payout system rests on two checks. The first is a list of approved payees, called beneficiaries. The second is a set of limits on how much can move. If someone can edit the list and get past the limits, they do not need to break the bank. The platform sends the money itself.
On October 9, 2026, CERT-In and CSIRT-Fin issued an advisory about attacks on payment and disbursement APIs. CERT-In is India's national cyber response team, and CSIRT-Fin is its counterpart for the financial sector. The agencies said attackers were abusing application weaknesses and taking API credentials. They said trusted systems were being used to route money to mule accounts, which are bank accounts that receive and pass on stolen funds. They also noted efforts to get around payee checks and transaction caps.
The advisory came less than a month before the reported PayU breach. MediaNama says investigators have not tied the attack to any weakness the advisory named. The two should not be treated as linked.
The arrests reach the cash-out end
The two men in custody, Atar Singh and Lovepreet Singh, are accused of providing bank accounts that received the stolen money. Police traced Rs 6.20 lakh to Atar Singh's account. In police's account, Atar Singh passed the account to Lovepreet Singh for a 2% commission, and Lovepreet then passed it to someone else for a further 2%.
That chain shows how cheaply the final step can be bought. It also leaves open the main questions. MediaNama reports that the investigation has not yet established how the attackers gained access, whether security controls failed, or whether merchant funds were affected.
Earlier cases, and a possible common thread
MediaNama lists earlier incidents. In September 2025, hackers allegedly breached a Hyderabad fintech firm and sent forged payment requests to Yes Bank, moving Rs 1.39 crore. In May 2022, Razorpay reported a Rs 7.38 crore fraud. Attackers had altered payment verification data, so 831 failed transactions were counted as successful.
These cases differ in method, and three incidents do not make a trend. This is our reading, not the source's: in the earlier two, money moved because a system trusted an instruction it should have checked. Whether that happened at PayU is not yet known.
What to ask your team
If your organisation runs automated payouts, put these questions to your security and payments leads. They come from this publication's argument that the payee list deserves as much protection as the login. The PayU case does not show what controls PayU had.
First, who or what can add a payee through the API, and does that change need a second approval from a separate system or a person? Second, where are API credentials stored, and are they limited to the task they serve? Third, what alert fires when a new payee appears, and who sees it at 11:30 pm?
Fourth, how long passes between a cloud provider flagging odd access and your own team acting on it? In the reported PayU case, the transfers were done before the unusual activity was seen. Last, have you checked your payout flows against the October 9 advisory?
A system that can name its own payees needs someone else to approve them.
Produced by the WebPulse Newsroom with AI assistance from the original reporting credited below, and checked against that source by our editorial review. How we use AI.
Original reporting: MediaNama.





