Nine in 10 Lenders Plan To Make Further Investments in AI Over the Next Two to Five Years

Artificial intelligence (AI) technology and its associated infrastructure is quickly becoming the preference of investors and lenders alike. As its use bleeds into every facet of fintech, a new report sheds light on how today’s attitudes towards AI are shaping its future. 

Brighterion, the AI and machine learning (ML) offshoot of Mastercard, has recently released its second annual report, ‘AI Perspectives: Credit Risk and Lending in collaboration with fintech media group LendIt.

Brighterion and LendIt collaborated on a survey sent to over 1,000 LendIt subscribers, including national and regional US banks, credit unions, community institutions, and financial technology providers.

The report advocates data-driven insights about financial institutions and lenders’ adoption of AI in credit risk management.

Amyn Dhala, chief product officer and VP of BrighterionAmyn Dhala, chief product officer and VP of Brighterion
Amyn Dhala

“Today’s report validates what we’re hearing from our customers with the continued and planned adoption of AI and its use cases, but there is still room to refine these approaches,” said Amyn Dhala, chief product officer and VP of Brighterion.

“Advanced credit risk management tools help to remove some of the guesswork and enable financial institutions to reduce losses while having more time to improve customer experience.”

Nine out of 10 respondents stated that they plan to make further investments in AI over the next two to five years.

Fifty-four per cent notes the ease of deploying the technology as a critical element to the success of the technology in the future.

According to the report, a further three in five lenders reported using AI for loan origination and 42 per cent use it for credit risk monitoring for existing customers. Seven in 10 respondents said they would like to use it for origination in the future, and three-quarters would like to use AI for credit risk monitoring for existing portfolios.

Seventy-nine per cent of respondents leveraged rules-based systems for credit risk management, with half reporting also using AI. The report recommends a combined approach that can layer in human expertise that better understands the context of the problem while leveraging the speed and processing of AI to reveal deeper insights.

Forty-six per cent said that model governance/regulatory requirements and hard to interpret ‘black-box’ models were top concerns for adopting AI for credit risk. As AI adoption grows, model explainability will become increasingly important and will result in new laws and regulations.

The report also identifies how lenders need to diversify their data sets to increase the breadth and depth of insights into their borrowers. Allegedly, credit bureau scores topped the list of data sources used to measure creditworthiness (80 per cent), followed by internal billing and payment history data (64 per cent).

  • Tyler is a Fintech Junior Journalist with specific interests in Online Banking and emerging AI technologies. He began his career writing with a plethora of national and international publications.