Artificial intelligence could transform how banks assess customers’ credit worthiness, according to the Governor of the Bank of England Mark Carney.
Speaking at the launch of Innovate Finance’s Global Summit in London on Monday (29 April), Carney identified credit assessments as one of the areas, alongside fraud analysis and cyber security, in which AI could have the biggest impact in the coming years.
“AI enabled solutions are increasingly important in fraud detection and automated threat intelligence and prevention,” he told delegates at a session marking the launch of London Fintech Week. “There’s also significant potential in cred ass, wholesale loan underwriting and trading.”
The role that algorithms play in making life-changing decisions, such as determining whether someone is eligible for a mortgage or business loan, has come under increased scrutiny in recent months as the so-called “techlash” has grown.
It is feared that companies are deploying AI and machine learning solutions which may amplify biases already existing in large datasets, leading to discriminatory decisions which are impossible to unpick or explain.
Carney also revealed that the Bank of England is using advanced analytics in order to assess the Prudential Regulation Authority’s rulebook, which is around 638,000 words long. “We want to find ways to simplify our rules and make compliance with them easier for firms,” he said.
“To explore ways to make reporting more efficient and effective were running a pilot with the Financial Conduct Authority on machine readable reporting requirements that firms systems can interpret and firms can automated regulatory data collection,” he added. “The initiatives are good in and of themselves but they are also ways of creating the potential to unlock the power of AI to do supervision better.”
The banking sector spends more on AI than any sector other than retail.