AI in banking applications for the front and middle office accounting could alone deliver $416 billion in savings.

Banks could save a humongous $447 billion by 2023 by deploying artificial intelligence (AI), says the AI IN BANKING research report from Business Insider Intelligence.

Of this, front office banking AI applications could yield $199 billion. However, the middle office and back office could save $217 billion and $31 billion respectively.

Meanwhile, according to a UBS Evidence Lab report viewed by Business Insider Intelligence, 75% of respondents at banks with over $100 billion in assets are already implementing AI strategies. For banks with less than $100 billion, that figure is 46%.

Meanwhile, banks have understood that AI can be used to harness powerful benefits for improving customer experience, reduction of costs, grow revenue, aid in compliance and penetrate new markets.

HSBC: Unbelievable opportunities

According to Josh Bottomley, HSBC’s global head of digital, artificial intelligence offers “unbelievable opportunities” in banking.

This is especially relevant in the context of cutting costs. Unfortunately, banks must now deal with cost pressures from low and negative interest rates globally. This trend has damaged their bread-and-butter business model of lending.

In an airline, you’ve still physically got to get a person from A to B, or a retailer, usually there is a good that’s there. However, the backend processes in banking are pretty much all data driven. “They’re all automatable, and they’re very susceptible for machine learning,” Bottomley says.

Where is AI most beneficial?

In the front office banks deploy artificial intelligence best for customer identification, improve customer relationships, in the use of chatbots and voice assistants, and in personalized banking, says the BI report.

In the middle-office, banks find artificial intelligence most useful for fraud identification and prevention and anti-money-laundering. It is also invaluable in know-your-customer (KYC) checks as required by regulators.

Further, AI also aids in credit risk assessment, pricing and underwriting.

What works for AI implementation in banks

According to the BI report, successful artificial intelligence implementation depends upon the foundation of a comprehensive AI strategy. This strategy should encompass data, external collaborations, and qualified and well-trained employees. It should cut across all the bank’s business lines.

[Related Story: Auriga Launches WWS AI For Advanced Banking Insights]