SecurityThreats & vulnerabilitiesNote
What CERT-In's 9 October advisory, The Economic Times' unnamed sources and Reuters reported, 9 to 30 October 2026, and what no regulator has confirmed.
Small non-bank lenders in India have lost about ₹300 crore to cyberattacks in roughly two months, The Economic Times reported on 29 October, citing unnamed “highly placed sources”.1 No regulator has published a loss figure: neither CERT-In’s advisories nor any RBI notice gives one.2 What is on the record is CERT-In’s “Critical” advisory of 9 October on attackers abusing payment and disbursement APIs,3 and, as Reuters reported on 30 October, an RBI push asking banks to prove their cyber controls work and to probe their API links with NBFCs and fintechs.4
The Economic Times puts the losses at about ₹300 crore over “the past two months or so”, with nearly 150 complaints or first information reports filed, mostly by smaller NBFCs. “Cyberattacks used to happen earlier also but for the past one and a half months, they have become rampant,” one unnamed person told the paper.1 MediaNama, reading the same report, describes the targets as lenders with turnover of about ₹1,000 crore or less, and one Delhi-based lender as losing ₹7.5 crore in one attack and ₹7.25 crore a few days later.2
Reuters adds one documented case. Two Mumbai-based non-bank lenders lost about $2 million through more than 700 unauthorised instant transfers from accounts at ICICI Bank on 8 and 9 October, according to a police complaint Reuters reviewed; the attackers had compromised the API connection that carries loan-disbursement instructions to the bank. Neither the RBI nor ICICI Bank responded immediately to Reuters.4
CERT-In’s advisory CIAD-2026-0046, “Threat Actors Exploiting Payment and Disbursement APIs for Unauthorized Fund Transfers in BFSI Entities”, names no victim and no loss. It describes the path, and says the time between compromise and fraudulent transfer has shortened, with possible use of AI-assisted tooling; it does not say AI was used in any given case.3
The loan system is skipped: a stolen credential calls the disbursal API directly
The attack path in CERT-In advisory CIAD-2026-0046, severity Critical, issued 9 October 2026
Break in at the edge
Internet-facing apps, abandoned APIs, hard-coded tokens, plaintext credentials and accounts without MFA.
Take the API keys
Web shells map the payment databases, then steal payment-gateway and partner-bank API credentials.
Call the disbursal API
From the lender’s own whitelisted servers, bypassing loan creation, KYC and sanction, above per-transaction limits.
Out through mule accounts
Rapid IMPS and NEFT transfers to mule accounts at several banks, cash withdrawn, logs deleted.
Source: CERT-In, CIAD-2026-0046 · as of 9 October 2026
Three people familiar with the discussions told Reuters the RBI raised its concerns in recent meetings with banks. It wants them to show that existing cybersecurity rules are implemented, to investigate weaknesses in their APIs, and to explain how they are preparing for AI-driven attacks.4
The reporting rules already exist, with a gap at the small end. The RBI’s NBFC Direction of 31 July 2026 requires reporting on its DAKSH platform “within six hours of detection” for Base Layer NBFCs with assets of ₹500 crore or more and for Middle, Upper and Top Layer NBFCs; its chapter for Base Layer NBFCs below ₹500 crore has no incident-reporting paragraph.5 Those smaller lenders still fall under CERT-In’s Directions of 28 April 2022, which require every body corporate to report listed incidents, including unauthorised access and attacks on digital payment systems, within six hours of noticing them.6
For a lender or fintech with a disbursement API, the advisory’s own controls are the checklist:3
The regulation behind each of these is mapped in TechBag’s BFSI guide; the credential controls are compared in the privileged access guide; and our earlier piece, two CERT-In advisories in six days, reads advisory 0046 alongside its companion.
Sources
Disclosure. TechBag is a software reseller: it earns a margin when a reader buys through it, including from vendors named here. The analysis is TechBag’s own; no vendor paid for or saw this piece before publication. Facts are labelled reported, TechBag count or our view.