CERT-In warns of payment workflow fraud and AI-enabled attacks
India’s CERT-In issued a critical advisory on November 20, 2026, covering abuse of business payment workflows and AI-driven agentic offensive frameworks. The reported warning describes payment fraud and AI-enabled attack tools, but does not establish that AI caused the payment fraud.
Key takeaways · 4
- 01
Review access to payment records and administrator accounts, as CERT-In recommends.
- 02
Independently verify payment rules and account-number changes before approval.
- 03
Monitor office logins and data leaving organisational networks.
- 04
CERT-In says cyberattacks on an organisation’s systems must be reported to the agency within six hours.
What CERT-In’s advisory covers
CERT-In issued advisory CIAD-2026-0062 on November 20, 2026, and rated it critical.[1] Its title covers abuse of business payment workflows alongside AI-driven agentic offensive frameworks.[1] The reported payment-fraud cases included organisations outside the banking and financial sector, including technology service providers supporting government payment processes and platforms handling vendor payments.[2] CERT-In also identified large companies, government offices, colleges and research centres as potential targets.[2] The advisory brings together concerns about payment workflows and AI-enabled tools; the evidence does not say the tools caused the payment fraud.[1][2]
How payment workflow fraud works
Rather than breaking through APIs or firewalls, attackers reportedly manipulated trusted business applications and payment systems.[2] CERT-In says they can change payment records or approval conditions, then transfer money through the genuine payment system.[2] Because the transaction uses that system, its records can appear valid and make the fraud hard to detect.[2] That makes verification of changes to payment rules and account numbers a practical control to prioritise: CERT-In specifically recommends checking those changes independently before approval.[2] Restricting access to payment records and key administrator accounts is another recommended step.[2]
What the AI warning describes
Separately, CERT-In’s reported warning describes tools that put several AI agents to work at the same time.[2] The agents are described as scanning systems for weaknesses, cracking passwords and gaining quiet network access.[2] CERT-In says the tools can adapt their approach when an attack hits a block or error, making the attacks harder to stop with fixed defences.[2] For security teams, that is a reason to review how they detect changing activity, alongside the agency’s advice to monitor office login activity and data leaving networks.[2]
Response steps and unanswered questions
CERT-In asked government departments, IT companies and online payment platforms to review their security.[2] It says organisations must report a cyberattack on their systems to the agency within six hours; suspected unauthorised withdrawals should also be reported through the cybercrime portal or by calling 1930.[2] Reuters reported that CERT-In did not say how much money had been improperly paid out or identify who received it.[3] Reuters also reported a recent spate of cyberattacks on banks and non-bank lenders in India, alongside warnings from authorities and tighter cybersecurity checks by the central bank.[3]
Payment fraud that uses legitimate workflows can leave records that look valid, so controls should cover approvals and changes to payment details—not just perimeter security. Teams should also treat the AI warning as a separate security concern and avoid assuming the advisory establishes a link between the two.
Why it matters
AI training for your team
kju.ai is built for teams: daily sessions, team analytics, and rollout support for your organization.
How this developed
22 November 2026
CERT-In warns of payment workflow fraud and AI-enabled attacks
Sources
- Advisoriescert-in.org.in
- CERT-In AI Alert: Hackers Target Payment Systems in India — The Open Presstheopenpress.in
- CERT-In flags unauthorised transactions, AI-driven cyberattacks in India - The HinduBusinessLinethehindubusinessline.com
