Thursday, November 05, 2026 | 10:37 AM ISTहिंदी में पढें
Business Standard
Notification Icon
userprofile IconSearch

Cert-In warns banks, fintech firms of AI-driven fund transfer fraud

India's cybersecurity watchdog has flagged an escalating AI-powered campaign targeting banks, NBFCs and fintech firms, with attackers exploiting payment APIs to carry out unauthorised fund transfers

digital infrastructure, Cyberattacks, Pahalgam attack, terrorist attacks, central government, Meity

Cert-In said some organisations had been deliberately targeted again, including those that had not fully addressed the vulnerabilities identified earlier.

Vrinda Goel New Delhi

Listen to This Article

Amid a rise in incidents of unauthorised fund transfers from bank accounts, India's cybersecurity watchdog, the Indian Computer Emergency Response Team (Cert-In), has issued a fresh advisory asking banking and financial services institutions (BFSI) to remain on high alert as the threat of cyber breaches continues.
 
In an advisory issued on Wednesday, Cert-In identified the campaign as "Crafty Falcon", an artificial intelligence (AI)-assisted cyber threat that uses publicly available AI tools to execute attacks autonomously and at machine speed. The tools enable attackers to generate and refine large numbers of customised attack scripts in quick succession.
 
The latest advisory is an update to an earlier warning issued on October 9 on "Threat Actors Exploiting Payment and Disbursement APIs for Unauthorised Fund Transfers in BFSI Entities".
  
According to Cert-In, continued analysis of incidents reported by affected organisations has revealed a significant escalation in the campaign. "The threat is active and ongoing. Computer Security Incident Response Team-Finance (CSIRT-Fin) continues to receive reports of successful attacks, including from NBFCs and fintech entities that had already received the earlier restricted advisories and had undertaken remediation," the advisory said.
 
Cert-In said some organisations had been deliberately targeted again, including those that had not fully addressed the vulnerabilities identified earlier. The attacks have also been timed to coincide with holidays, weekends and other periods of reduced security monitoring, delaying their detection and response.
 

What tools are attackers using to target financial institutions?

 
Cert-In's analysis has identified several publicly available AI-driven offensive security tools being used against affected organisations. These include the ARTEX family of large language model (LLM)-based tools, the PentAGI autonomous multi-agent AI penetration testing system, the Cairn attack search engine and the CyberStrikeAI framework.
 
According to the advisory, these tools can execute entire attack chains autonomously, generate and refine customised attack scripts in rapid succession and adapt their payloads based on how target systems respond.
 

How are attackers targeting the financial institutions?

 
Cert-In has identified several methods used by attackers to target financial institutions. These include exploiting internet-facing testing and staging environments to gain access to production systems through shared credentials, tokens and network connections.
 
Attackers have also targeted cloud infrastructure, abusing services such as AWS Systems Manager and Lambda to execute commands and deploy malicious files while blending their activities with legitimate operations. In other cases, they have reverse-engineered application code and used modified application packages to initiate unauthorised API calls.
 
The advisory also flags attempts to bypass IP-based security controls by routing requests through whitelisted third-party services. Attackers have used micro-value transactions, sometimes as low as ₹1, to test unauthorised fund transfers before moving larger sums. Merchant settlement APIs of payment aggregators have also been exploited to move funds.
 
Cert-In has further reported the exploitation of newly disclosed vulnerabilities, including file-upload and deserialisation weaknesses in PHP and Node.js applications. Cryptocurrency exchanges and virtual digital asset service providers have also been targeted, with attackers stealing wallet keys and siphoning funds.
 

What steps should banks and fintech firms take to stay safe?

 
Cert-In has advised BFSI entities to strengthen round-the-clock security monitoring, with additional staffing and lower alert thresholds during weekends, holidays and other periods of reduced coverage. It has also called for real-time monitoring of cloud audit logs, tighter controls on cloud management services and restrictions on serverless functions, including disabling internet access by default.
 
Financial institutions should isolate testing and staging environments from production systems, avoid sharing credentials and API keys, and strengthen security monitoring of internet-facing applications. The watchdog has also recommended server-side validation of transactions, API request signing using nonces and timestamps, and app attestation to detect altered or repackaged applications.
 
Entities have been asked to establish faster patching timelines for internet-facing applications, particularly those with file-upload and deserialisation vulnerabilities. Cert-In has also advised replacing IP-based access controls for shared third-party services with identity-based safeguards.
 
On the transaction side, institutions should flag small-value transfers to new beneficiaries followed by high-value transactions, treating such patterns as potential indicators of fraud. Cryptocurrency platforms have been advised to secure wallet keys through hardware security modules or cold custody, require multi-signature approvals for withdrawals and monitor unusual withdrawal activity.
 
Cert-In has further recommended restricting unauthorised AI-driven security and attack tools on corporate systems and monitoring LLM API usage. Organisations have also been advised to adopt zero-trust principles, micro-segmentation and identity-based access controls.
 
In the event of a suspected breach, affected organisations should isolate compromised systems, preserve forensic evidence and logs, rotate credentials and report the incident to Cert-In or CSIRT-Fin.
 

Don't miss the most important news and views of the day. Get them on our Telegram channel

First Published: Nov 05 2026 | 10:34 AM IST