CERT-In Warns AI Attacks Now Targeting Govt Payouts Too

CERT-In warns AI-driven attacks are now abusing business payment workflows to divert funds from banks, fintech firms and government payout systems.

Soniya Mishra
7 Min Read

India’s cyber security agency has issued its sharpest warning yet about a new kind of attack, one that uses artificial intelligence to move money out of companies and government payment systems at machine speed. The Indian Computer Emergency Response Team, CERT-In, said in a critical advisory on November 20 that hackers are no longer just targeting banks. They are now going after the business payment workflows behind everyday payouts, from vendor settlements to government-linked disbursements.

The warning matters because of who is in the firing line. It is not only banks and fintech firms any more. Service providers that handle payments for government departments, companies that pay vendors and partners, and platforms that process bulk payouts are all now named as targets.

What CERT-In said on November 20

The advisory says CERT-In has received reports of unauthorised financial transactions involving organisations outside the financial sector. Some incidents hit service providers that support government-linked payment processes and platforms that handle vendor or partner payouts.

The agency’s central point is a shift in method. Earlier attacks mostly went after payment and disbursement APIs, the technical interfaces that move money. The new attacks go after the business records and approval steps that decide whether a payment should happen at all. Attackers alter those records, so a transaction that should never have been approved sails through the usual checks.

How the attacks work

  • Attackers first get in through unpatched software or internet-facing servers that have been wrongly configured, including customer-facing systems and complaint-handling interfaces.
  • Once inside, they target automated fund disbursement systems and the interfaces used to process payments and transfers.
  • Before moving large sums, they run tiny test transactions, sometimes as small as Re 1, to confirm that unauthorised transfers will go through.
  • They exploit merchant settlement APIs belonging to payment aggregators to move the funds, and in crypto cases they steal wallet keys and drain wallets.

Individual cases have involved unauthorised transactions worth several crores, and financial institutions have been told to report incidents within six hours of detection.

The AI tools behind the attacks

What makes this campaign different is speed and scale. CERT-In says the attackers use AI-driven agentic offensive frameworks, tools that can run complete attack chains on their own, carrying out reconnaissance, finding exposed systems and preparing exploits, all in parallel and at machine speed.

The advisory names several publicly available tools: the ARTEX family of LLM-based penetration testing tools built on the PentAGI autonomous multi-agent framework, the CyberStrikeAI framework, and the Cairn attack search engine. The attackers also exploit newly published vulnerabilities as soon as details become public, and generate large numbers of tailored attack scripts in quick succession, a pattern consistent with AI-generated code.

A security analyst monitoring payment system alerts in a control room
Security teams are being told to watch the business records behind payments, not just the payment interfaces. (Representative image)

From banks to government payouts

The November 20 advisory is the third step in a rapid escalation. On October 9, CERT-In issued advisory CIAD-2026-0046 about attackers exploiting payment and disbursement APIs. On November 4, advisory CIAD-2026-0058 identified a previously undocumented AI-assisted threat actor, named CRAFTY FALCON, and warned that the campaign was active and still escalating.

The earlier advisories named non-banking financial companies, digital lending platforms, fintech firms, payment aggregators, payment service providers and cryptocurrency exchanges as victims. The latest one adds technology service providers supporting government-linked payment processes and platforms handling vendor or partner payouts.

A pattern, not a one-off

The warning landed in the same week as a separate, confirmed breach at a major Indian software company. Customer engagement platform MoEngage confirmed on November 21 that it suffered a security breach in late October after a compromised third-party vendor account gave attackers access to a limited part of its cloud infrastructure. Fewer than one per cent of its customer accounts were affected, the company said, but crypto exchange ZebPay told its users that information about some of them had been taken, including contact details and recent account activity. ZebPay said its own servers, wallets and trading systems were not affected and customer funds remain safe.

Why it matters

For ordinary readers, the direct risk is not that a bank vault gets emptied. It is that the systems businesses and government departments use to pay each other can be quietly turned against them, and the effects eventually reach customers through disrupted services and stolen personal details that fuel impersonation scams.

For companies, the lesson is that securing the payment API is no longer enough. The records and approvals that feed into a payment need the same protection, with checks that verify the business data behind each transaction rather than trusting it because it arrived through a familiar system.

Frequently asked questions

What is CERT-In?
CERT-In is India’s national agency for responding to cyber security incidents, working under the Ministry of Electronics and Information Technology.

What is new about the November 20 advisory?
Earlier warnings focused on attacks against payment APIs in banks and fintech firms. The new advisory says attackers are now manipulating the business payment workflows and approval records behind those payments, and that government-linked payment processes are also being targeted.

Who is CRAFTY FALCON?
It is the name CERT-In gave to a previously undocumented, AI-assisted threat actor identified in its November 4 advisory, which it said was attacking banks, lenders and payment firms.

India’s cyber defenders are now racing tools that work at machine speed. The November 20 advisory is a clear signal that the defence has to move just as fast.

Ask about this articleInstant answers from this articleAI
AI-generated answers. Ask about the article, or anything else.
Share This Article
Soniya Mishra covers technology, world news and politics for Informeia, with over 7 years of experience in news reporting. She tracks tech launches, global developments and political stories for Informeia readers. She holds a degree in journalism and mass communication.
Leave a Comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Choose language