New Delhi: Use of artificial intelligence and machine learning can transform credit delivery to India’s vast majority of poor and strengthen financial inclusion by providing credit services to over 41 crore Jan Dhan account holders in addition to basic saving facilities, says a former official of the Reserve Bank of India.
S Ganesh Kumar, former executive director in the department of payment and settlement systems (DPSS) of the Reserve Bank of India, says banks have not been using artificial intelligence and machine learning in credit delivery as use of AI is still in its early days.
Kumar says banks and their banking correspondents (BCs) know that a Jan Dhan account has been opened for a customer and only financial transactions and delivery of subsidies through direct benefit transfer (DBT) are taking place but no credit or no loan is being given to these customers.
“The bank system can have the list of people who have basic savings accounts but do not have credit facilities. It can be matched with Aadhaar and can be traced about the usage of money that is coming,” Kumar told the audience in the Business and Banking dialogue organised by Mumbai based fintech firm EPS India.
Over 41 crore Jan-Dhan accounts
According to the latest official data, there are over 41.31 crore Jan Dhan (basic banking services) accounts in the country which also have a deposit of over Rs 1.31 lakh crores. The Union government uses these Aadhaar linked bank accounts for direct transfer of cash subsidy to farmers, students, senior citizens, widows, poor and disabled persons among others.
The Centre also uses these Aadhaar linked bank accounts for transfer of Prime Minister Kisan Samman Nidhi to a vast majority of the country’s over 12.5 crore small and marginal farmers.
Under the PM Kisan Samman Nidhi, the Centre transfers Rs 6,000 per year to eligible farmers in three installments of Rs 2,000 each for buying input materials for agriculture such as seeds and fertilisers among other things.
S Ganesh Kumar, who supervised the digital payments and settlement systems in the country during his days with India’s banking sector regulator, says creditworthiness of these account holders can be assessed by analysing the transaction history of their bank accounts.
“If he got Rs 2,000 under the DBT, it can be found out how he used this money, Rs 500 for seeds, Rs 1000 for fertilisers and Rs 100, perhaps for drinking. If the government can furnish this information then it can be found out whether this person is credit worthy or not,” Kumar told the audience in response to a question by ETV Bharat.
AI can help in matching bank profile, customer requirement
He says this information can be accessed by a nearby bank and after matching the profile of the bank and customer’s requirement a loan can be offered to the account holder.
“For example, a loan of Rs 1,000 can be given to him with the condition that if he doesn't repay then this Rs 1,000 amount can be deducted from the next DBT installment that he receives,” explained the banker.
“Here is a loan customer, who was in the financial system only for basic banking services but now he becomes a customer for credit delivery. Once he becomes a customer, he spreads the word around and the message reaches to others,” he added.
Kumar, however, warns that use of artificial intelligence and machine learning in delivery of credit services in the country is still in its infancy and may entail some risks like non-payment by some customers.
“Everything has a positive and negative side. We will have to fine tune it but this can be done,” he said.
(Article by Krishnanand Tripathi)
