RBI aims to drive adoption of AI tool among banks

RBIH chief executive officer Rajesh Bansal said the AI ​​platform, called Mulehunter.Ai, has already been implemented by two large public sector banks so far.

Bansal added that the accuracy rate of the new AI tool is three times that of the banks’ previous manual systems.

The AI ​​tool was developed after identifying 19 patterns, including sudden transactions in an inactive account, multiple credits in an account followed by a large debit, among others

The innovation arm of the Reserve Bank of India (RBI) is reportedly in talks with 10 public and private sector banks to accelerate the adoption of its artificial intelligence (AI)-based platform to detect cases of financial fraud through “mulle accounts “.

Soft accounts refer to bank accounts used for illegal activities such as money laundering, fraud and other illegal operations.

Reserve Bank Innovation Hub (RBIH) executive director Rajesh Bansal told the Economic Times that the AI ​​platform, called Mulehunter.Ai, has already been implemented by two large public sector banks. He added that early results from implementing the tool were “encouraging in terms of much better accuracy and reduced time.”

“If the percent accuracy of the manual system was X, this is 3X. If the time taken was Y, this is 0.1 Y,” Bansal said.

The RBIH chief executive said the AI ​​tool was developed after identifying 19 patterns, including sudden transactions in a dormant account, multiple credits in an account followed by a large debit, among others.

These patterns were identified after RBIH conducted stakeholder consultations with as many as 10 banks to ascertain the traditional approach adopted by lenders while flagging soft accounts.

Insights gathered from the meeting showed that most banks were still using a rules-based system. According to this, suspicious accounts were first identified and then the verification was done manually in each case, which resulted in significant loss of time.

However, the new AI platform was intended to speed up the detection of fraudulent accounts. “The best way would be to see where the money ends up – in the mule accounts,” Bansal said, adding that there are multiple channels through which fraud can occur and the amount involved can be as high as INR 1 Cr “in these days” .

The comments come months after RBI Governor Shaktikanta Das in July this year urged public and private sector banks to strengthen their security measures to crack the whip on mule accounts and digital fraud.

Later in October, reports emerged that the central bank was looking to develop an AI-enabled system to alert people in real time to financial fraud and enable banks and financial institutions to detect soft accounts.

According to government data, Indians lost as much as INR 11,333 crore to cyber fraud in the first nine months of 2024 alone. Just a few days ago, Minister of State for Finance Pankaj Chaudhary informed Parliament that the country recorded 6.32 lakh UPI fraud cases worth INR 485 Cr in the first six months of the current financial year 2024-25 (FY25)