Fintech platforms have emerged as an unlikely collateral damage as Yes Bank strengthens its defences against AI-driven cyberattacks.

Sources told Inc42 that the cybersecurity upgrades at Yes Bank are disrupting workflows and delaying payouts at payment service providers. 

One fintech executive said that his platform was facing issues ever since Yes Bank sent an email, which noted that the bank was “implementing enhanced precautionary measures to safeguard” transactions. He was subsequently also informed that the bank was making changes to its banking APIs, which was delaying settlements.

These changes, in turn, are impacting payment processing and settlement timelines for some businesses that rely on banking APIs to move funds.

Another industry insider noted that the issues began to crop up after the Reserve Bank of India (RBI) issued directives to banks and other financial institutions to bolster their cybersecurity defences. According to the source quoted above, the RBI directions came after the recent cyberattack on the State Bank of India (SBI), referring to the data leak at the bank in October.

The RBI is also said to have directed the banks and other financial institutions to ensure the funds of customers are not impacted by the growing wave of AI-assisted cyberattacks.

Inc42 has reached out to Yes Bank for a comment on the matter. This story will be updated upon receiving their response.

This comes as multiple Indian crypto exchanges continue to grapple with delays in processing withdrawals due to disruptions at its banking partners. Coinbase India had paused all withdrawals last week, before partially resuming payouts, citing an outage at its banking partner.

CoinDCX, which has also been facing similar issues, claimed that upgrades at “most” of its partner banks was delaying the processing of INR withdrawals. Zebpay also said that payouts were taking longer than usual because of a technical disruption at its banking partner.

Earlier yesterday, Razorpay also told its customers that it had to delay merchant settlements temporarily due to a technical issue at a partner bank. However, the IPO-bound unicorn later reinstated the service hours later.

The development comes on the same day as CERT-In has issued an alarm over unauthorised transactions at technology service providers for banks and other financial institutions. 

Bank Vendors Under AI Attack

In an advisory dated November 21, the cybersecurity watchdog flagged that malicious actors were using AI-driven cyberattacks to undertake unauthorised transactions at technology service providers, which handle government-linked payment processes, as well as platforms handling vendor payouts.

CERT-In said that attackers altered business records or approval relationships and ​then used legitimate payment processes to approve fraudulent transactions. The watchdog did not elaborate on how much money of the proceeds had been funneled out. 

The cybersecurity agency said that the reported incidents involved cyberattackers misusing business applications and workflows, rather than direct exploitation of payment APIs alone. As per CERT-In, these attackers used trusted systems to process unauthorised transactions and  exploited gaps in verification and reconciliation of changes to business records.

CERT-In also noted that it has identified the following techniques used by attackers for perpetrating AI-assisted cyberattacks against fintech platforms:

  • Automated reconnaissance to identify exposed systems, services and possible attack paths
  • Rapid preparation and modification of exploits for known vulnerabilities, including through analysis of software updates and related technical information
  • Use of multiple agents or automated workers to undertake reconnaissance, privilege discovery and other intrusion tasks in parallel
  • Modification of scripts, requests and attack steps in response to errors, blocked attempts or information returned by targeted systems
  • Use of valid credentials, legitimate administrative functions, trust relationships and external infrastructure to gain further access.

As such, CERT-In also directed government entities to identify and restrict accounts and integrations that can approve payments or change payment-related records. 

Besides recommending undertaking independent certification for high-risk changes to payment records, it also called on the banks to reconcile critical business data maintained by service providers against authoritative records. 

It also called on the TSPS to restrict administrative accounts and control-plane privileges. CERT-In also directed fintechs to review unusual activity and monitor changes to payment-related business records. 

The post Fintechs Bear The Brunt As Yes Bank Upgrades AI Safeguards appeared first on Inc42 Media.